public ITransformer Train(IDataView trainingData)
{
TrainedModel = _trainingPipeline.Fit(trainingData);
PredictionFunction = TrainedModel.MakePredictionFunction <DemandObservation, DemandPrediction>(_mlcontext);
return(TrainedModel);
}
public static void Example()
{
// Create a new ML context, for ML.NET operations. It can be used for exception tracking and logging,
// as well as the source of randomness.
var mlContext = new MLContext();
var samples = new List <DataPoint>()
{
new DataPoint()
{
Label = 3, Features = new float[3] {
1, 1, 0
}
},
new DataPoint()
{
Label = 32, Features = new float[3] {
0, float.NaN, 1
}
},
new DataPoint()
{
Label = float.NaN, Features = new float[3] {
-1, float.NaN, -3
}
},
};
// Convert training data to IDataView, the general data type used in ML.NET.
var data = mlContext.Data.LoadFromEnumerable(samples);
// IndicateMissingValues is used to create a boolean containing
// 'true' where the value in the input column is NaN. This value can be used
// to replace missing values with other values.
IEstimator <ITransformer> pipeline = mlContext.Transforms.IndicateMissingValues("MissingIndicator", "Features");
// Now we can transform the data and look at the output to confirm the behavior of the estimator.
// This operation doesn't actually evaluate data until we read the data below.
var tansformer = pipeline.Fit(data);
var transformedData = tansformer.Transform(data);
// We can extract the newly created column as an IEnumerable of SampleDataTransformed, the class we define below.
var rowEnumerable = mlContext.Data.CreateEnumerable <SampleDataTransformed>(transformedData, reuseRowObject: false);
// a small printing utility
Func <object[], string> vectorPrinter = (object[] vector) =>
{
string preview = "[";
foreach (var slot in vector)
{
preview += $"{slot} ";
}
return(preview += "]");
};
// And finally, we can write out the rows of the dataset, looking at the columns of interest.
foreach (var row in rowEnumerable)
{
Console.WriteLine($"Label: {row.Label} Features: {vectorPrinter(row.Features.Cast<object>().ToArray())} MissingIndicator: {vectorPrinter(row.MissingIndicator.Cast<object>().ToArray())}");
}
// Expected output:
//
// Label: 3 Features: [1 1 0] MissingIndicator: [False False False]
// Label: 32 Features: [0 NaN 1] MissingIndicator: [False True False]
// Label: NaN Features: [-1 NaN -3 ] MissingIndicator: [False True False]
}
public void IrisLightGbm()
{
if (!RuntimeInformation.IsOSPlatform(OSPlatform.Windows))
{
// https://github.com/dotnet/machinelearning/issues/4156
return;
}
var mlContext = new MLContext(seed: 1);
var connectionString = GetConnectionString(TestDatasets.irisDb.name);
var commandText = $@"SELECT * FROM ""{TestDatasets.irisDb.trainFilename}""";
var loaderColumns = new DatabaseLoader.Column[]
{
new DatabaseLoader.Column()
{
Name = "Label", Type = DbType.Int32
},
new DatabaseLoader.Column()
{
Name = "SepalLength", Type = DbType.Single
},
new DatabaseLoader.Column()
{
Name = "SepalWidth", Type = DbType.Single
},
new DatabaseLoader.Column()
{
Name = "PetalLength", Type = DbType.Single
},
new DatabaseLoader.Column()
{
Name = "PetalWidth", Type = DbType.Single
}
};
var loader = mlContext.Data.CreateDatabaseLoader(loaderColumns);
var databaseSource = new DatabaseSource(SqlClientFactory.Instance, connectionString, commandText);
var trainingData = loader.Load(databaseSource);
IEstimator <ITransformer> pipeline = mlContext.Transforms.Conversion.MapValueToKey("Label")
.Append(mlContext.Transforms.Concatenate("Features", "SepalLength", "SepalWidth", "PetalLength", "PetalWidth"))
.Append(mlContext.MulticlassClassification.Trainers.LightGbm())
.Append(mlContext.Transforms.Conversion.MapKeyToValue("PredictedLabel"));
var model = pipeline.Fit(trainingData);
var engine = mlContext.Model.CreatePredictionEngine <IrisData, IrisPrediction>(model);
Assert.Equal(0, engine.Predict(new IrisData()
{
SepalLength = 4.5f,
SepalWidth = 5.6f,
PetalLength = 0.5f,
PetalWidth = 0.5f,
}).PredictedLabel);
Assert.Equal(1, engine.Predict(new IrisData()
{
SepalLength = 4.9f,
SepalWidth = 2.4f,
PetalLength = 3.3f,
PetalWidth = 1.0f,
}).PredictedLabel);
}
public void TrainModel(string testImagePath = null)
{
#region Notes: Fundamental components
/* Main components:
* IDataView,
* ITransformer,
* IEstimator
*/
//IDataView demoDataView;
//ITransformer demoITransformer;
//IEstimator<ITransformer> demoIEstimator;
#endregion Notes: Fundamental components
#region Notes: Conventional column names
/* Conventional column names:
* Input:
* Label
* Features
* Output:
* PredictedLabel
* Score
*/
#endregion Notes: Conventional column names
#region Notes: Usual training process
/* Usual training process:
* 1. Load training/test datasets (IDataView)
* 2. Build training pipeline (IEstimator)
* 2.1 Construct preProcessing pipeline (IEstimator) (optional)
* 2.2 Configure trainer (IEstimator)
* 2.3 Construct postProcessing pipeline (optional)
* 2.4 Construct training pipeline (preProcessing pipelin + trainer + postProcessing pipline
* 3. Train model using training dataset (ITransformer)
* 4. Evaluate model perfomance
* 4.1 Make predictions on test data using trained model (IDataView)
* 4.2 Compute evaluation metrics (Metrics staticsitcs)
* (optional) Retrain on full dataset (Itransformer)
* 5. Save model to filesystem
* 6. Make single prediction
*/
#endregion Notes: Usual training process
// Load data
IDataView imagesInfo = LoadData(_dataFolder);
imagesInfo = mlContext.Data.ShuffleRows(imagesInfo);
DataOperationsCatalog.TrainTestData dataSplit = mlContext.Data.TrainTestSplit(imagesInfo, testFraction: 0.2);
// Pre processing
IEstimator <ITransformer> e_preProcessing_readImageBytes = mlContext.Transforms.LoadRawImageBytes(
inputColumnName: nameof(ImageFileInputModel.ImagePath),
outputColumnName: nameof(ImageInputModel.Image),
imageFolder: _dataFolder);
IEstimator <ITransformer> e_preProcessing_labelKeyMapping = mlContext.Transforms.Conversion.MapValueToKey(
inputColumnName: nameof(BaseInputModel.Label),
outputColumnName: "LabelAsKey",
keyOrdinality: Microsoft.ML.Transforms.ValueToKeyMappingEstimator.KeyOrdinality.ByValue);
ITransformer t_preProcessing_labelKeyMapping = e_preProcessing_labelKeyMapping.Fit(imagesInfo);
ITransformer t_preProcessing_readImageBytes = e_preProcessing_readImageBytes.Fit(imagesInfo);
ITransformer t_preProcessingPipeline = t_preProcessing_labelKeyMapping.Append(t_preProcessing_readImageBytes);
// Core Model training pipeline
IDataView testSetTransformed = t_preProcessingPipeline.Transform(dataSplit.TestSet);
ImageClassificationTrainer.Options trainerSettings = new ImageClassificationTrainer.Options
{
FeatureColumnName = nameof(ImageInputModel.Image),
LabelColumnName = "LabelAsKey",
Arch = ImageClassificationTrainer.Architecture.ResnetV2101,
Epoch = 100,
BatchSize = 200,
LearningRate = 0.05f,
MetricsCallback = (m) => Console.WriteLine(m),
ValidationSet = testSetTransformed,
WorkspacePath = _workspaceFolder
};
IEstimator <ITransformer> e_trainer = mlContext.MulticlassClassification.Trainers.ImageClassification(trainerSettings);
IEstimator <ITransformer> e_postProcessing_labelKeyMapping = mlContext.Transforms.Conversion.MapKeyToValue(
inputColumnName: "PredictedLabel",
outputColumnName: nameof(PredictionModel.PredictedLabel));
IEstimator <ITransformer> trainingPipeline = e_trainer.Append(e_postProcessing_labelKeyMapping);
// Train
#region Notes: On metadata
/*
* Metadata source: https://aka.ms/mlnet-resources/resnet_v2_101_299.meta
* System.IO.Path.GetTempPath() - C:\Users\User\AppData\Local\Temp\
*/
#endregion
ITransformer trainedModel = Train(trainingPipeline, t_preProcessingPipeline.Transform(dataSplit.TrainSet));
#region Notes: Model composition
//var extractPixelsEst = mlContext.Transforms.ExtractPixels();
//var resizeEst = mlContext.Transforms.ResizeImages();
//IEstimator<ITransformer> est = mlContext.Model.LoadTensorFlowModel("MODEL_PATH")
//.ScoreTensorFlowModel(
//outputColumnNames: new[] { "some-name" },
//inputColumnNames: new[] { "Features" }, addBatchDimensionInput: true);
#endregion Model composition
// Evaluate/Save FileSystemModel
ITransformer fileSystemModel = t_preProcessingPipeline.Append(trainedModel);
Evaluate(fileSystemModel, dataSplit.TestSet);
SaveModel(fileSystemModel,
new DataViewSchema.Column[] {
imagesInfo.Schema.First(x => x.Name == nameof(ImageFileInputModel.ImagePath)),
imagesInfo.Schema.First(x => x.Name == nameof(BaseInputModel.Label))
},
ResolveModelFileName("fromFile"));
// Evaluate/Save InMemoryModel
IDataView testSetImageExtracted = t_preProcessing_readImageBytes.Transform(dataSplit.TrainSet);
ITransformer inMemoryModel = t_preProcessing_labelKeyMapping.Append(trainedModel);
Evaluate(inMemoryModel, testSetImageExtracted);
SaveModel(inMemoryModel,
new DataViewSchema.Column[] {
testSetImageExtracted.Schema.First(x => x.Name == nameof(ImageFileInputModel.ImagePath)),
testSetImageExtracted.Schema.First(x => x.Name == nameof(BaseInputModel.Label))
},
ResolveModelFileName("inMemory"));
//Try a single prediction
if (!string.IsNullOrWhiteSpace(testImagePath))
{
MakeSinglePrediction(testImagePath);
}
}
private ITransformer TrainModel(MLContext mlContext, IDataView trainingDataView, IEstimator <ITransformer> trainingPipeline)
{
ITransformer model = trainingPipeline.Fit(trainingDataView);
return(model);
}
protected virtual ITransformer BuildAndTrainModel(IDataView trainingData, IEstimator <ITransformer> pipeline)
{
return(pipeline.Fit(trainingData));
}
static void Main(string[] args)
{
Helper.PrintLine("创建 MLContext...");
MLContext mlContext = new MLContext(seed: 0);
ITransformer model;
IDataView testDataView = mlContext.Data.LoadFromTextFile <MovieRating>(TestingDataPath, hasHeader: true, separatorChar: ',');
if (File.Exists(ModelPath))
{
Helper.PrintLine("加载神经网络模型...");
model = mlContext.Model.Load(ModelPath, out DataViewSchema inputScema);
}
else
{
// 数据集合
IDataView trainingDataView = mlContext.Data.LoadFromTextFile <MovieRating>(TrainingDataPath, hasHeader: true, separatorChar: ',');
// 创建神经网络管道
Helper.PrintLine("创建神经网络管道...");
IEstimator <ITransformer> estimator = mlContext.Transforms.Conversion
.MapValueToKey(outputColumnName: "userIdEncoded", inputColumnName: "userId")
.Append(mlContext.Transforms.Conversion.MapValueToKey(outputColumnName: "movieIdEncoded", inputColumnName: "movieId"))
.Append(mlContext.Recommendation().Trainers.MatrixFactorization(
new MatrixFactorizationTrainer.Options
{
MatrixColumnIndexColumnName = "userIdEncoded",
MatrixRowIndexColumnName = "movieIdEncoded",
LabelColumnName = "Label",
NumberOfIterations = 20,
ApproximationRank = 100
}));
// 开始训练神经网络
Helper.PrintSplit();
Helper.PrintLine("开始训练神经网络...");
model = estimator.Fit(trainingDataView);
Helper.PrintLine("训练神经网络完成");
Helper.PrintSplit();
Helper.PrintLine($"导出神经网络模型...");
mlContext.Model.Save(model, trainingDataView.Schema, ModelPath);
}
// 预测
Helper.PrintLine("预测：");
var prediction = model.Transform(testDataView);
var metrics = mlContext.Regression.Evaluate(prediction, labelColumnName: "Label", scoreColumnName: "Score");
Helper.PrintLine($"R^2: {metrics.RSquared:0.##}");
Helper.PrintLine($"RMS error: {metrics.RootMeanSquaredError:0.##}");
var predictionEngine = mlContext.Model.CreatePredictionEngine <MovieRating, MovieRatingPrediction>(model);
var testInput = new MovieRating {
userId = 6, movieId = 10
};
var movieRatingPrediction = predictionEngine.Predict(testInput);
if (Math.Round(movieRatingPrediction.Score, 1) > 3.5)
{
Helper.PrintLine($"Movie {testInput.movieId} is recommended for user {testInput.userId}");
}
else
{
Helper.PrintLine($"Movie {testInput.movieId} is not recommended for user {testInput.userId}");
}
Helper.Exit(0);
}
public void Train(string trainingFileName, string testFileName)
{
if (!File.Exists(trainingFileName))
{
Console.WriteLine($"Failed to find training data file ({trainingFileName}");
return;
}
if (!File.Exists(testFileName))
{
Console.WriteLine($"Failed to find test data file ({testFileName}");
return;
}
var trainingDataView = MlContext.Data.LoadFromTextFile <CarInventory>(trainingFileName, ',', hasHeader: false);
IEstimator <ITransformer> dataProcessPipeline = MlContext.Transforms.Concatenate("Features",
typeof(CarInventory).ToPropertyList <CarInventory>(nameof(CarInventory.Label)))
.Append(MlContext.Transforms.NormalizeMeanVariance(inputColumnName: "Features",
outputColumnName: "FeaturesNormalizedByMeanVar"));
var dataSplit = MlContext.Data.TrainTestSplit(trainingDataView, testFraction: 0.2);
var trainer = MlContext.BinaryClassification.Trainers.FastTree(labelColumnName: nameof(CarInventory.Label),
featureColumnName: "FeaturesNormalizedByMeanVar",
numberOfLeaves: 2,
numberOfTrees: 1000,
minimumExampleCountPerLeaf: 1,
learningRate: 0.2);
var trainingPipeline = dataProcessPipeline.Append(trainer);
var trainedModel = trainingPipeline.Fit(trainingDataView);
MlContext.Model.Save(trainedModel, trainingDataView.Schema, ModelPath);
var evaluationPipeline = trainedModel.Append(MlContext.Transforms
.CalculateFeatureContribution(trainedModel.LastTransformer)
.Fit(dataProcessPipeline.Fit(trainingDataView).Transform(trainingDataView)));
var testDataView = MlContext.Data.LoadFromTextFile <CarInventory>(testFileName, ',', hasHeader: false);
var testSetTransform = evaluationPipeline.Transform(testDataView);
var modelMetrics = MlContext.BinaryClassification.Evaluate(data: testSetTransform,
labelColumnName: nameof(CarInventory.Label),
scoreColumnName: "Score");
Console.WriteLine($"Accuracy: {modelMetrics.Accuracy:P2}");
Console.WriteLine($"Area Under Curve: {modelMetrics.AreaUnderRocCurve:P2}");
Console.WriteLine($"Area under Precision recall Curve: {modelMetrics.AreaUnderPrecisionRecallCurve:P2}");
Console.WriteLine($"F1Score: {modelMetrics.F1Score:P2}");
Console.WriteLine($"LogLoss: {modelMetrics.LogLoss:#.##}");
Console.WriteLine($"LogLossReduction: {modelMetrics.LogLossReduction:#.##}");
Console.WriteLine($"PositivePrecision: {modelMetrics.PositivePrecision:#.##}");
Console.WriteLine($"PositiveRecall: {modelMetrics.PositiveRecall:#.##}");
Console.WriteLine($"NegativePrecision: {modelMetrics.NegativePrecision:#.##}");
Console.WriteLine($"NegativeRecall: {modelMetrics.NegativeRecall:P2}");
}
static void Main(string[] args)
{
//###############################################################
//INICIALIZACIÓN DEL PROCESO
//###############################################################
//Inicialización de mlContext; utilización del seed para replicidad
MLContext mlContext = new MLContext(seed: 1);
//Definición de las clases de los datos de entrada:
// -Clase Observaciones: TransactionObservation
//Carga de datos
IDataView originalFullData = mlContext.Data.LoadFromTextFile <TransactionObservation>(
_DataPath,
separatorChar: ';',
hasHeader: true);
//###############################################################
//CONSTRUYE EL CONJUNTO DE DATOS (DATASET)
//###############################################################
//División del IDataView originalFullData:
// -entrenamiento (trainingDataView): 70%
// -testeo (testDataView): 20%
// -Consumo (ConsumoDataView): 10%
//Split dataset: train = 0.7 + test_Consumo = 0.3
double testFraction = 0.3;
TrainTestData Split_TrainTestConsumoData = mlContext.Data.TrainTestSplit(originalFullData,
testFraction: testFraction, seed: 1);
IDataView trainingDataView = Split_TrainTestConsumoData.TrainSet;
IDataView testConsumoData = Split_TrainTestConsumoData.TestSet;
//Split dataset tes_val: test = 0.7 (0.7*0.3 = 0.21) + val = 0.3 (0.3*0.3 = 0.09)
testFraction = 0.3;
TrainTestData Split_TestConsumoData = mlContext.Data.TrainTestSplit(testConsumoData,
testFraction: testFraction, seed: 1);
IDataView testDataView = Split_TestConsumoData.TrainSet;
IDataView ConsumoDataView = Split_TestConsumoData.TestSet;
//save train split
using (var fileStream = File.Create(_salida_trainDataPath))
{
mlContext.Data.SaveAsText(trainingDataView, fileStream, separatorChar: ';', headerRow: true,
schema: true);
}
//save test split
using (var fileStream = File.Create(_salida_testDataPath))
{
mlContext.Data.SaveAsText(testDataView, fileStream, separatorChar: ';', headerRow: true,
schema: true);
}
//save Consumo split
using (var fileStream = File.Create(_salida_ConsumoDataPath))
{
mlContext.Data.SaveAsText(ConsumoDataView, fileStream, separatorChar: ';', headerRow: true,
schema: true);
}
//###############################################################
//SELECCIÓN DE VARIABLES
//###############################################################
//Suprimimos del esquema IDataView lo que no seleccionemos como features
string[] featureColumnNames = trainingDataView.Schema.AsQueryable()
.Select(column => column.Name)
.Where(name => name != "Label" && //atributo de salida
name != "Time") //no aporta información
.ToArray();
//###############################################################
//TRANFORMACIÓN DE LOS DATOS DEL MODELO --> pipeline
//###############################################################
//Concatena
IEstimator <ITransformer> pipeline = mlContext.Transforms.Concatenate("Features",
featureColumnNames)
//Surpime del IDataView
.Append(mlContext.Transforms.DropColumns(new string[] { "Time" }))
//Normalizado de las Features
.Append(mlContext.Transforms.NormalizeMeanVariance(inputColumnName: "Features",
outputColumnName: "FeaturesNormalized"));
//Guardar dataset transformedData --> Validación Cruzada
IDataView transformedData =
pipeline.Fit(trainingDataView).Transform(trainingDataView);
using (var fileStream = File.Create(_salida_transformationData))
{
mlContext.Data.SaveAsText(transformedData, fileStream, separatorChar: ';', headerRow: true,
schema: true);
}
//###############################################################
//SELECCIÓN DE ALGORITMOS DE ENTRENAMIENTO --> trainingPipeline
//###############################################################
//***************************************************************
//1. SVM (Suport Vector Machine)
//***************************************************************
var trainer_svm = mlContext.BinaryClassification.Trainers
.LinearSvm(labelColumnName: "Label",
featureColumnName: "FeaturesNormalized",
numberOfIterations: 10);
//Se añade el Algoritmo al pipeline de transformación de datos
IEstimator <ITransformer> trainingPipeline_svm = pipeline.Append(trainer_svm);
//***************************************************************
//2. GBA (Gradient Boosting Algorithm)
//***************************************************************
var trainer_boost = mlContext.BinaryClassification.Trainers
.FastTree(labelColumnName: "Label",
featureColumnName: "FeaturesNormalized",
numberOfLeaves: 20,
numberOfTrees: 100,
minimumExampleCountPerLeaf: 10,
learningRate: 0.2);
//Se añade el Algoritmo al pipeline de transformación de datos
IEstimator <ITransformer> trainingPipeline_boost = pipeline.Append(trainer_boost);
//###############################################################
//ENTRENAMIENTO DE LOS MODELOS
//###############################################################
Console.WriteLine($"\n************************************************************");
Console.WriteLine($"* Entrenamiento del Modelo calculado con el Algoritmo SVM ");
Console.WriteLine($"*-----------------------------------------------------------");
var watch_svm = System.Diagnostics.Stopwatch.StartNew();
var model_svm = trainingPipeline_svm.Fit(trainingDataView);
watch_svm.Stop();
var elapseds_svm = watch_svm.ElapsedMilliseconds * 0.001;
Console.WriteLine($"El entrenamiento SVM ha tardado: {elapseds_svm:#.##} s\n");
Console.WriteLine($"\n************************************************************");
Console.WriteLine($"* Entrenamiento del Modelo calculado con el Algoritmo GBA ");
Console.WriteLine($"*-----------------------------------------------------------");
var watch_boost = System.Diagnostics.Stopwatch.StartNew();
var model_boost = trainingPipeline_boost.Fit(trainingDataView);
watch_boost.Stop();
var elapseds_boost = watch_boost.ElapsedMilliseconds * 0.001;
Console.WriteLine($"El entrenamiento GBA ha tardado: {elapseds_boost:#.##} s\n");
//###############################################################
//EVALUACIÓN DE LOS MODELOS
//###############################################################
//Transformación del IDataView testDataView a paritr de ambos Modelos
var predictions_svm = model_svm.Transform(testDataView);
var predictions_boost = model_boost.Transform(testDataView);
//Calculo de las métricas de cada Modelo
var metrics_svm = mlContext.BinaryClassification
//SVM es un Modelo no basado en PROBABILIDAD -> NonCalibrated
.EvaluateNonCalibrated(data: predictions_svm, labelColumnName: "Label", scoreColumnName: "Score");
var metrics_boost = mlContext.BinaryClassification
.Evaluate(data: predictions_boost, labelColumnName: "Label", scoreColumnName: "Score");
//Muestra las métricas SVM
Console.WriteLine($"\n************************************************************");
Console.WriteLine($"* Métricas para el Modelo calculado con el Algoritmo SVM ");
Console.WriteLine($"*-----------------------------------------------------------");
Console.WriteLine($"* SVM Positive Precision: {metrics_svm.PositivePrecision:0.##}");
Console.WriteLine($"* SVM Positive Recall: {metrics_svm.PositiveRecall:0.##}");
Console.WriteLine($"* SVM Negative Precision: {metrics_svm.NegativePrecision:0.##}");
Console.WriteLine($"* SVM Negative Recall: {metrics_svm.NegativeRecall:0.##}");
Console.WriteLine($"* SVM Accuracy: {metrics_svm.Accuracy:P2}");
Console.WriteLine($"* SVM F1Score: {metrics_svm.F1Score:P2}\n");
//Muestra las métricas GBA
Console.WriteLine($"\n************************************************************");
Console.WriteLine($"* Métricas para el Modelo calculado con el Algoritmo GBA ");
Console.WriteLine($"*-----------------------------------------------------------");
Console.WriteLine($"* GBA Positive Precision: {metrics_boost.PositivePrecision:0.##}");
Console.WriteLine($"* GBA Positive Recall: {metrics_boost.PositiveRecall:0.##}");
Console.WriteLine($"* GBA Negative Precision: {metrics_boost.NegativePrecision:0.##}");
Console.WriteLine($"* GBA Negative Recall: {metrics_boost.NegativeRecall:0.##}");
Console.WriteLine($"* GBA Accuracy: {metrics_boost.Accuracy:P2}");
Console.WriteLine($"* GBA F1Score: {metrics_boost.F1Score:P2}\n");
//###############################################################
//VALIDACIÓN CRUZADA
//###############################################################
Console.WriteLine($"\n*****************************************");
Console.WriteLine($"* Validación Cruzada del Algoritmo SVM ");
Console.WriteLine($"*----------------------------------------");
var watch_CV_SVM = System.Diagnostics.Stopwatch.StartNew();
var cvResults_svm = mlContext.BinaryClassification
//SVM es un Modelo no basado en PROBABILIDAD -> NonCalibrated
.CrossValidateNonCalibrated(
transformedData,
trainer_svm,
numberOfFolds: 10,
labelColumnName: "Label");
watch_CV_SVM.Stop();
var elapseds_CV_SVM = watch_CV_SVM.ElapsedMilliseconds * 0.001;
Console.WriteLine($"La Validación Cruzada del Algoritmo SVM ha tardado: {elapseds_CV_SVM:#.##} s\n");
//Vamos a supervisar el resultado de la Validación Cruzada para la métrica: F1 Score
Double[] F1_models =
cvResults_svm
.OrderByDescending(fold => fold.Metrics.F1Score)
.Select(fold => fold.Metrics.F1Score)
.ToArray();
//Calculamos la media del F1 Score
Double media_F1 = F1_models.Average();
//Vamos a supervisar el resultado de la Validación Cruzada para la métrica: Accuracy
Double[] Accu_models =
cvResults_svm
.OrderByDescending(fold => fold.Metrics.F1Score)
.Select(fold => fold.Metrics.Accuracy)
.ToArray();
//Calculamos la media del Accuracy
Double media_Accu = Accu_models.Average();
//Mostramos métricas y media
Console.WriteLine($"\n**********************************************************");
Console.WriteLine($"* Resultado de la Validación Cruzada del Algoritmo SVM ");
Console.WriteLine($"*---------------------------------------------------------");
Console.WriteLine($"| MODEL_N | MEDIDA F1 SCORE | MEDIDA ACCURACY |");
Console.WriteLine($"| Model_1 | {F1_models[0]:P2} | {Accu_models[0]:P2} |");
Console.WriteLine($"| Model_2 | {F1_models[1]:P2} | {Accu_models[1]:P2} |");
Console.WriteLine($"| Model_3 | {F1_models[2]:P2} | {Accu_models[2]:P2} |");
Console.WriteLine($"| Model_4 | {F1_models[3]:P2} | {Accu_models[3]:P2} |");
Console.WriteLine($"| Model_5 | {F1_models[4]:P2} | {Accu_models[4]:P2} |");
Console.WriteLine($"| Model_6 | {F1_models[5]:P2} | {Accu_models[5]:P2} |");
Console.WriteLine($"| Model_7 | {F1_models[6]:P2} | {Accu_models[6]:P2} |");
Console.WriteLine($"| Model_8 | {F1_models[7]:P2} | {Accu_models[7]:P2} |");
Console.WriteLine($"| Model_9 | {F1_models[8]:P2} | {Accu_models[8]:P2} |");
Console.WriteLine($"| Model_10 | {F1_models[9]:P2} | {Accu_models[9]:P2} |");
Console.WriteLine($"La F1 Score media es igual a: {media_F1:P2}");
Console.WriteLine($"La Accuracy media es igual a: {media_Accu:P2}\n");
//###############################################################
//SELECCIÓN MODELO
//###############################################################
//Tomamos todos los Modelos calculados con la Validación Cruzada
ITransformer[] models =
cvResults_svm
.OrderByDescending(fold => fold.Metrics.F1Score)
.Select(fold => fold.Model)
.ToArray();
//Tomamos el mejor Modelo
ITransformer topModel = models[0];
//Guardamos el Modelo para su posterior consumo
mlContext.Model.Save(model_svm, trainingDataView.Schema, _salida_modelPath);
//######################################
//CONSUMO DEL MODELO
//######################################
//Definición de las clases de las predicciones:
// -Clase Predicciones: TransactionPrediction
//Definimos CreatePredictionEngine de TransactionObservation --> TransactionPrediction a través de model_svm
var predictionEngine = mlContext.Model
.CreatePredictionEngine <TransactionObservation, TransactionPrediction>(
model_svm);
Console.WriteLine($"\n**********************************");
Console.WriteLine($"--- Predicción ConsumoDataView ---");
Console.WriteLine($"----------------------------------");
mlContext.Data.CreateEnumerable <TransactionObservation>(ConsumoDataView, reuseRowObject: false)
.Select(ConsumoData => ConsumoData)
.ToList()
.ForEach(ConsumoData =>
{
//Predict() predicción única instancia
var prediction = predictionEngine.Predict(ConsumoData);
Console.WriteLine($"Label: {prediction.Label:.##}");
Console.WriteLine($"Predicted Label: {prediction.PredictedLabel:#.##}");
//SVM no está basado en determinar Probabilidad
//Console.WriteLine("Probability: {prediction.Probability:#.##}");
Console.WriteLine($"Score: {prediction.Score:.##}");
Console.WriteLine($"-------------------");
});
}
//calibrated version below
public static IReadOnlyList <TrainCatalogBase.CrossValidationResult <CalibratedBinaryClassificationMetrics> > BuildAndTrain(MLContext mlContext, IDataView splitTrainSet)
//public static IReadOnlyList<TrainCatalogBase.CrossValidationResult<BinaryClassificationMetrics>> BuildAndTrain(MLContext mlContext, IDataView splitTrainSet)
{
//var estimator = mlContext.Transforms.Text.FeaturizeText(outputColumnName: "Features", inputColumnName: nameof(QuestionData.QuestionText))
//transformedData used to be pipeline when algorithm was in it also
IEstimator <ITransformer> transformedDataStage1 = mlContext.Transforms.Text.FeaturizeText(inputColumnName: "UserAbility", outputColumnName: "UserAbilityFeaturized")
.Append(mlContext.Transforms.Text.FeaturizeText(inputColumnName: "QuestionDifficulty", outputColumnName: "QuestionDifficultyFeaturized")
.Append(mlContext.Transforms.Concatenate("Features", "UserAbilityFeaturized", "QuestionDifficultyFeaturized")));
//used below to help with stop watch
//https://docs.microsoft.com/en-us/dotnet/api/system.diagnostics.stopwatch?redirectedfrom=MSDN&view=netframework-4.8
Console.WriteLine("===================== Starting Stopwatch ====================");
Stopwatch stopWatch = new Stopwatch();
stopWatch.Start();
//what does the fit method do? These two methods involve transforming the data for use in the algorithm, unspecific though
var dataPrepTransformer = transformedDataStage1.Fit(splitTrainSet);
transformedDataStage2 = dataPrepTransformer.Transform(splitTrainSet);
Console.WriteLine(splitTrainSet.Schema.ToString());
//LogReg Stopchastic used as this was given in the sample, which used calibrated model, will change to SVM
IEstimator <ITransformer> svmLinAlg = mlContext.BinaryClassification.Trainers.Prior();
var cvResults = mlContext.BinaryClassification.CrossValidate(transformedDataStage2, svmLinAlg, numberOfFolds: 10);
//svm linear now used
//IEstimator<ITransformer> svmLinAlg = mlContext.BinaryClassification.Trainers.FastForest();
//var cvResults = mlContext.BinaryClassification.CrossValidateNonCalibrated(transformedDataStage2, svmLinAlg, numberOfFolds: 10);
//Console.WriteLine("=============== Create and Train the Model ===============");
//Console.WriteLine("=============== End of training ===============");
//Console.WriteLine();
//the cvResults object will contain a lot of things:
// 1. TrainTestData object for each fold of data
// 2. a model for each fold
// 3. a metric for each fold
Console.WriteLine("===================== Stopwatch Stopped=========================");
stopWatch.Stop();
TimeSpan ts = stopWatch.Elapsed;
string elapsedTime = String.Format("{0:00}:{1:00}:{2:00}.{3:00}",
ts.Hours, ts.Minutes, ts.Seconds,
ts.Milliseconds / 10);
Console.WriteLine($"Time taken (hh:mm:ss:msms): {elapsedTime}");
return(cvResults);
}
static void Main()
{
MLContext mlContext = new MLContext();
// 1. Uvezi ili stvori trening podatke
HouseData[] houseData =
{
new HouseData {
Size = 1.1F, Price = 1.2F
},
new HouseData {
Size = 1.9F, Price = 2.3F
},
new HouseData {
Size = 2.8F, Price = 3.0F
},
new HouseData {
Size = 3.4F, Price = 3.7F
}
};
IDataView trainingData = mlContext.Data.LoadFromEnumerable(houseData);
// 2. Specificiraj pipeline za pripremu podataka i trening
IEstimator <ITransformer> pipeline = mlContext
.Transforms
.Concatenate(
outputColumnName: "Features",
inputColumnNames: new[] { "Size" })
.Append(mlContext.Regression.Trainers.Sdca(
labelColumnName: "Price",
featureColumnName: "Features",
maximumNumberOfIterations: 100));
// 3. Treniraj model
ITransformer model = pipeline.Fit(trainingData);
// 4. Testiraj model
HouseData[] testHouseData =
{
new HouseData {
Size = 1.1F, Price = 0.98F
},
new HouseData {
Size = 1.9F, Price = 2.1F
},
new HouseData {
Size = 2.8F, Price = 2.9F
},
new HouseData {
Size = 3.4F, Price = 3.6F
}
};
IDataView testHouseDataView = mlContext.Data.LoadFromEnumerable(testHouseData);
IDataView testPriceDataView = model.Transform(testHouseDataView);
var debug = testPriceDataView.Preview();
var metrics = mlContext.Regression.Evaluate(
testPriceDataView,
labelColumnName: "Price");
Console.WriteLine($"R^2: {metrics.RSquared:0.##}");
Console.WriteLine($"RMS error: {metrics.RootMeanSquaredError:0.##}");
// 5. Spremi model
mlContext.Model.Save(model, trainingData.Schema, "model.zip");
// 6. Radi predikciju
MakePrediction();
}
public List <FeatureImportance> ComputePermutationMetrics(string trainingDataPath)
{
IEstimator <ITransformer> pipeline =
MLContext.Transforms.ReplaceMissingValues(
outputColumnName: "FixedAcidity",
replacementMode: MissingValueReplacingEstimator.ReplacementMode.Mean)
.Append(MLContext.Transforms.Concatenate("Features",
new[]
{
"FixedAcidity",
"VolatileAcidity",
"CitricAcid",
"ResidualSugar",
"Chlorides",
"FreeSulfurDioxide",
"TotalSulfurDioxide",
"Density",
"Ph",
"Sulphates",
"Alcohol"
}))
.Append(MLContext.Transforms.NormalizeMeanVariance("Features"));
var trainData = MLContext.Data.LoadFromTextFile <FeatureImportanceData>(
path: trainingDataPath,
separatorChar: ';',
hasHeader: true);
// Cache the data view in memory. For an iterative algorithm such as SDCA this makes a huge difference.
trainData = MLContext.Data.Cache(trainData);
var transformationModel = pipeline.Fit(trainData);
// Prepare the data for the algorithm.
var transformedData = transformationModel.Transform(trainData);
// Choose a regression algorithm.
var algorithm = MLContext.Regression.Trainers.Sdca();
// Train the model and score it on the transformed data.
var regressionModel = algorithm.Fit(transformedData);
// Calculate the PFI metrics.
var permutationMetrics = MLContext.Regression.PermutationFeatureImportance(
regressionModel,
transformedData,
permutationCount: 50);
// List of evaluation metrics:
// https://docs.microsoft.com/en-us/dotnet/machine-learning/resources/metrics
var result = new List <FeatureImportance> {
new FeatureImportance("FixedAcidity"),
new FeatureImportance("VolatileAcidity"),
new FeatureImportance("CitricAcid"),
new FeatureImportance("ResidualSugar"),
new FeatureImportance("Chlorides"),
new FeatureImportance("FreeSulfurDioxide"),
new FeatureImportance("TotalSulfurDioxide"),
new FeatureImportance("Density"),
new FeatureImportance("Ph"),
new FeatureImportance("Sulphates"),
new FeatureImportance("Alcohol")
};
for (int i = 0; i < permutationMetrics.Length; i++)
{
result[i].R2Decrease = permutationMetrics[i].RSquared.Mean;
}
return(result);
}
public IEnumerable <RatioAnalysis> Predict(IEnumerable <RatioAnalysis> data)
{
var mlContext = new MLContext(seed: 1);
IEstimator <ITransformer> costofgoodsForcaster = mlContext.Forecasting.ForecastBySsa(
outputColumnName: nameof(RatioAnalysisPrediction.Forecasted),
inputColumnName: nameof(RatioAnalysis.CostOfGoods), // This is the column being forecasted.
windowSize: 12, // Window size is set to the time period represented in the product data cycle; our product cycle is based on 12 months, so this is set to a factor of 12, e.g. 3.
seriesLength: data.Count(), // This parameter specifies the number of data points that are used when performing a forecast.
trainSize: data.Count(), // This parameter specifies the total number of data points in the input time series, starting from the beginning.
horizon: 3, // Indicates the number of values to forecast; 3 indicates that the next 3 months of product units will be forecasted.
confidenceLevel: 0.75f, // Indicates the likelihood the real observed value will fall within the specified interval bounds.
confidenceLowerBoundColumn: nameof(RatioAnalysisPrediction.ConfidenceLowerBound), //This is the name of the column that will be used to store the lower interval bound for each forecasted value.
confidenceUpperBoundColumn: nameof(RatioAnalysisPrediction.ConfidenceUpperBound)); //This is the name of the column that will be used to store the upper interval bound for each forecasted value.
IEstimator <ITransformer> inventoryForcaster = mlContext.Forecasting.ForecastBySsa(
outputColumnName: nameof(RatioAnalysisPrediction.Forecasted),
inputColumnName: nameof(RatioAnalysis.Inventory), // This is the column being forecasted.
windowSize: 12, // Window size is set to the time period represented in the product data cycle; our product cycle is based on 12 months, so this is set to a factor of 12, e.g. 3.
seriesLength: data.Count(), // This parameter specifies the number of data points that are used when performing a forecast.
trainSize: data.Count(), // This parameter specifies the total number of data points in the input time series, starting from the beginning.
horizon: 3, // Indicates the number of values to forecast; 3 indicates that the next 3 months of product units will be forecasted.
confidenceLevel: 0.75f, // Indicates the likelihood the real observed value will fall within the specified interval bounds.
confidenceLowerBoundColumn: nameof(RatioAnalysisPrediction.ConfidenceLowerBound), //This is the name of the column that will be used to store the lower interval bound for each forecasted value.
confidenceUpperBoundColumn: nameof(RatioAnalysisPrediction.ConfidenceUpperBound)); //This is the name of the column that will be used to store the upper interval bound for each forecasted value.
IEstimator <ITransformer> turnoverForcaster = mlContext.Forecasting.ForecastBySsa(
outputColumnName: nameof(RatioAnalysisPrediction.Forecasted),
inputColumnName: nameof(RatioAnalysis.Turnover), // This is the column being forecasted.
windowSize: 12, // Window size is set to the time period represented in the product data cycle; our product cycle is based on 12 months, so this is set to a factor of 12, e.g. 3.
seriesLength: data.Count(), // This parameter specifies the number of data points that are used when performing a forecast.
trainSize: data.Count(), // This parameter specifies the total number of data points in the input time series, starting from the beginning.
horizon: 3, // Indicates the number of values to forecast; 3 indicates that the next 3 months of product units will be forecasted.
confidenceLevel: 0.75f, // Indicates the likelihood the real observed value will fall within the specified interval bounds.
confidenceLowerBoundColumn: nameof(RatioAnalysisPrediction.ConfidenceLowerBound), //This is the name of the column that will be used to store the lower interval bound for each forecasted value.
confidenceUpperBoundColumn: nameof(RatioAnalysisPrediction.ConfidenceUpperBound)); //This is the name of the column that will be used to store the upper interval bound for each forecasted value.
// Fit the forecasting model to the specified product's data series.
ITransformer costofgoodsTransformer = costofgoodsForcaster.Fit(mlContext.Data.LoadFromEnumerable(data));
ITransformer inventoryTransformer = inventoryForcaster.Fit(mlContext.Data.LoadFromEnumerable(data));
ITransformer turnoverTransformer = turnoverForcaster.Fit(mlContext.Data.LoadFromEnumerable(data));
// Create the forecast engine used for creating predictions.
TimeSeriesPredictionEngine <RatioAnalysis, RatioAnalysisPrediction> inventoryEngine = inventoryTransformer.CreateTimeSeriesEngine <RatioAnalysis, RatioAnalysisPrediction>(mlContext);
TimeSeriesPredictionEngine <RatioAnalysis, RatioAnalysisPrediction> turneroverEngine = turnoverTransformer.CreateTimeSeriesEngine <RatioAnalysis, RatioAnalysisPrediction>(mlContext);
TimeSeriesPredictionEngine <RatioAnalysis, RatioAnalysisPrediction> costofgoodEngine = costofgoodsTransformer.CreateTimeSeriesEngine <RatioAnalysis, RatioAnalysisPrediction>(mlContext);
// Get the prediction; this will include the forecasted turnover for the next 3 months since this
//the time period specified in the `horizon` parameter when the forecast estimator was originally created.
var turnoverPrediction = turneroverEngine.Predict();
var costPrediction = costofgoodEngine.Predict();
var inventoryPrediction = inventoryEngine.Predict();
var last = data.Last();
var retVal = data.ToList();
for (int i = 0; i < turnoverPrediction.Forecasted.Count(); i++)
{
retVal.Add(new RatioAnalysis
{
Date = last.Date.AddMonths(i + 1),
CostOfGoods = costPrediction.Forecasted[i],
CostOfGoodsDelta = costPrediction.ConfidenceUpperBound[i] - costPrediction.Forecasted[i],
Inventory = inventoryPrediction.Forecasted[i],
InventoryDelta = inventoryPrediction.ConfidenceUpperBound[i] - inventoryPrediction.Forecasted[i],
Turnover = turnoverPrediction.Forecasted[i],
TurnoverDelta = turnoverPrediction.ConfidenceUpperBound[i] - turnoverPrediction.Forecasted[i]
});
}
return(retVal);
}
/// <summary>
/// Fits the training data to the training pipeline.
/// </summary>
/// <returns> The model </returns>
private static ITransformer TrainModel()
{
return(TrainingPipeline.Fit(TrainingDataView));
}
public ITransformer TrainFeaturizeText()
{
var textColumns = new List <string>();
for (int i = 0; i < 20; i++) // Only load first 20 columns
{
textColumns.Add($"Column{i}");
}
var featurizers = new List <TextFeaturizingEstimator>();
foreach (var textColumn in textColumns)
{
var featurizer = _mlContext.Transforms.Text.FeaturizeText(textColumn, new TextFeaturizingEstimator.Options()
{
CharFeatureExtractor = null,
WordFeatureExtractor = new WordBagEstimator.Options()
{
NgramLength = 2,
MaximumNgramsCount = new int[] { 200000 }
}
});
featurizers.Add(featurizer);
}
IEstimator <ITransformer> pipeline = featurizers.First();
foreach (var featurizer in featurizers.Skip(1))
{
pipeline = pipeline.Append(featurizer);
}
var model = pipeline.Fit(_dataset);
// BENCHMARK OUTPUT
// * Summary *
//BenchmarkDotNet = v0.11.3, OS = Windows 10.0.18363
//Intel Xeon W - 2133 CPU 3.60GHz, 1 CPU, 12 logical and 6 physical cores
//.NET Core SDK = 3.0.100
//[Host] : .NET Core 2.1.13(CoreCLR 4.6.28008.01, CoreFX 4.6.28008.01), 64bit RyuJIT
//Job - KDKCUJ : .NET Core 2.1.13(CoreCLR 4.6.28008.01, CoreFX 4.6.28008.01), 64bit RyuJIT
//Arguments =/ p:Configuration = Release Toolchain = netcoreapp2.1 IterationCount = 1
//LaunchCount = 3 MaxIterationCount = 20 RunStrategy = ColdStart
//UnrollFactor = 1 WarmupCount = 1
// Method | Mean | Error | StdDev | Extra Metric | Gen 0 / 1k Op | Gen 1 / 1k Op | Gen 2 / 1k Op | Allocated Memory / Op |
//------------------- | --------:| --------:| ---------:| -------------:| -------------:| ------------: | ------------: | --------------------: |
// TrainFeaturizeText | 17.00 s | 6.337 s | 0.3474 s | - | 1949000.0000 | 721000.0000 | 36000.0000 | 315.48 MB |
//// * Legends *
// Mean : Arithmetic mean of all measurements
// Error : Half of 99.9 % confidence interval
// StdDev : Standard deviation of all measurements
// Extra Metric: Value of the provided extra metric
// Gen 0 / 1k Op : GC Generation 0 collects per 1k Operations
// Gen 1 / 1k Op : GC Generation 1 collects per 1k Operations
// Gen 2 / 1k Op : GC Generation 2 collects per 1k Operations
// Allocated Memory/ Op : Allocated memory per single operation(managed only, inclusive, 1KB = 1024B)
// 1 s: 1 Second(1 sec)
//// * Diagnostic Output - MemoryDiagnoser *
//// ***** BenchmarkRunner: End *****
// Run time: 00:01:52(112.92 sec), executed benchmarks: 1
//// * Artifacts cleanup *
// Global total time: 00:01:59(119.89 sec), executed benchmarks: 1
return(model);
}
/// <summary>
/// Fits the training data to the training pipeline.
/// </summary>
/// <returns> model </returns>
public static ITransformer TrainModel()
{
ITransformer model = TrainingPipeline.Fit(TrainingDataView);
return(model);
}
// Build and train model
public static ITransformer GenerateModel(MLContext mlContext)
{
// <SnippetImageTransforms>
IEstimator <ITransformer> pipeline = mlContext.Transforms.LoadImages(outputColumnName: "input", imageFolder: _imagesFolder, inputColumnName: nameof(ImageData.ImagePath))
// The image transforms transform the images into the model's expected format.
.Append(mlContext.Transforms.ResizeImages(outputColumnName: "input", imageWidth: InceptionSettings.ImageWidth, imageHeight: InceptionSettings.ImageHeight, inputColumnName: "input"))
.Append(mlContext.Transforms.ExtractPixels(outputColumnName: "input", interleavePixelColors: InceptionSettings.ChannelsLast, offsetImage: InceptionSettings.Mean))
// </SnippetImageTransforms>
// The ScoreTensorFlowModel transform scores the TensorFlow model and allows communication
// <SnippetScoreTensorFlowModel>
.Append(mlContext.Model.LoadTensorFlowModel(_inceptionTensorFlowModel).
ScoreTensorFlowModel(outputColumnNames: new[] { "softmax2_pre_activation" }, inputColumnNames: new[] { "input" }, addBatchDimensionInput: true))
// </SnippetScoreTensorFlowModel>
// <SnippetMapValueToKey>
.Append(mlContext.Transforms.Conversion.MapValueToKey(outputColumnName: "LabelKey", inputColumnName: "Label"))
// </SnippetMapValueToKey>
// <SnippetAddTrainer>
.Append(mlContext.MulticlassClassification.Trainers.LbfgsMaximumEntropy(labelColumnName: "LabelKey", featureColumnName: "softmax2_pre_activation"))
// </SnippetAddTrainer>
// <SnippetMapKeyToValue>
.Append(mlContext.Transforms.Conversion.MapKeyToValue("PredictedLabelValue", "PredictedLabel"))
.AppendCacheCheckpoint(mlContext);
// </SnippetMapKeyToValue>
// <SnippetLoadData>
IDataView trainingData = mlContext.Data.LoadFromTextFile <ImageData>(path: _trainTagsTsv, hasHeader: false);
// </SnippetLoadData>
// Train the model
Console.WriteLine("=============== Training classification model ===============");
// Create and train the model
// <SnippetTrainModel>
ITransformer model = pipeline.Fit(trainingData);
// </SnippetTrainModel>
// Generate predictions from the test data, to be evaluated
// <SnippetLoadAndTransformTestData>
IDataView testData = mlContext.Data.LoadFromTextFile <ImageData>(path: _testTagsTsv, hasHeader: false);
IDataView predictions = model.Transform(testData);
// Create an IEnumerable for the predictions for displaying results
IEnumerable <ImagePrediction> imagePredictionData = mlContext.Data.CreateEnumerable <ImagePrediction>(predictions, true);
DisplayResults(imagePredictionData);
// </SnippetLoadAndTransformTestData>
// Get performance metrics on the model using training data
Console.WriteLine("=============== Classification metrics ===============");
// <SnippetEvaluate>
MulticlassClassificationMetrics metrics =
mlContext.MulticlassClassification.Evaluate(predictions,
labelColumnName: "LabelKey",
predictedLabelColumnName: "PredictedLabel");
// </SnippetEvaluate>
//<SnippetDisplayMetrics>
Console.WriteLine($"LogLoss is: {metrics.LogLoss}");
Console.WriteLine($"PerClassLogLoss is: {String.Join(" , ", metrics.PerClassLogLoss.Select(c => c.ToString()))}");
//</SnippetDisplayMetrics>
// <SnippetReturnModel>
return(model);
// </SnippetReturnModel>
}
static void Main(string[] args)
{
//###############################################################
//INICIALIZACIÓN DEL PROCESO
//###############################################################
//Inicialización de mlContext; utilización del seed para replicidad
MLContext mlContext = new MLContext(seed: 1);
//Definición de las clases de los datos de entrada:
// -Clase Observaciones: BorderCrossObservation
//Carga de datos
IDataView originalFullData =
mlContext.Data.LoadFromTextFile <BorderCrossObservation>(
_DataPath,
separatorChar: ';',
hasHeader: true);
//###############################################################
//CONSTRUYE EL CONJUNTO DE DATOS (DATASET)
//###############################################################
//División del IDataView originalFullData:
// -entrenamiento (trainingDataView): 80%
// -testeo (testDataView): 20%
//Selección de porcentaje para el test
double testFraction = 0.2;
//Aplicacón de la División
TrainTestData Split_trainTestData = mlContext.Data.TrainTestSplit(originalFullData,
testFraction: testFraction, seed: 1);
//IDataView resultantes
IDataView trainingDataView = Split_trainTestData.TrainSet;
IDataView testDataView = Split_trainTestData.TestSet;
//Guardar IDataView trainingDataView para una posible viasualización (extensión csv)
using (var fileStream = File.Create(_salida_trainDataPath))
{
mlContext.Data.SaveAsText(trainingDataView, fileStream, separatorChar: ';', headerRow: true, schema: true);
}
//Guardar IDataView testDataView para una posible viasualización (extensión csv)
using (var fileStream = File.Create(_salida_testDataPath))
{
mlContext.Data.SaveAsText(testDataView, fileStream, separatorChar: ';', headerRow: true, schema: true);
}
//###############################################################
//SELECCIÓN DE VARIABLES
//###############################################################
//Suprimimos del esquema IDataView lo que no seleccionemos como features
var listfeatureColumnNames = trainingDataView.Schema.AsQueryable()
.Select(column => column.Name)
.Where(name => name != "Label" && //atributo de salida
name != "Port_Name" && //solo existe un valor
name != "State" && //un valor
name != "Port_Code" && //un valor
name != "Border" && //un valor
name != "Longitud" && //un valor
name != "Latitud" && //un valor
name != "Mes" && //transformar
name != "Year" && //transformar
name != "Measure").ToList(); //transformar
//Añadimos las Transformaciones de los atributos suprimidos anteriormente
listfeatureColumnNames.Add("MesInd");
listfeatureColumnNames.Add("YearInd");
listfeatureColumnNames.Add("MeasureInd");
//Conversión a array para su posterior utlización
string[] featureColumnNames = listfeatureColumnNames.ToArray();
//###############################################################
//TRANFORMACIÓN DE LOS DATOS DEL MODELO --> pipeline
//###############################################################
//Indicadoras
IEstimator <ITransformer> pipeline = mlContext.Transforms.Categorical.OneHotEncoding(
outputColumnName: "MesInd", inputColumnName: "Mes")
//Indicadoras
.Append(mlContext.Transforms.Categorical.OneHotEncoding(
outputColumnName: "YearInd", inputColumnName: "Year"))
//Indicadoras
.Append(mlContext.Transforms.Categorical.OneHotEncoding(
outputColumnName: "MeasureInd", inputColumnName: "Measure"))
//Concatena
.Append(mlContext.Transforms.Concatenate(
"Features", featureColumnNames))
//Surpime del IDataView
.Append(mlContext.Transforms.DropColumns(
new string[] { "Mes", "Year", "Measure" }))
//Normalizado del atributo de salida
.Append(mlContext.Transforms.NormalizeMeanVariance(
inputColumnName: "Label", outputColumnName: "LabelNormalized"));
//Guardar datos transformedData
IDataView transformedData =
pipeline.Fit(trainingDataView).Transform(trainingDataView);
using (var fileStream = File.Create(_salida_transformationData))
{
mlContext.Data.SaveAsText(transformedData, fileStream, separatorChar: ';', headerRow: true, schema: true);
}
//###############################################################
//SELECCIÓN DE ALGORITMOS DE ENTRENAMIENTO --> trainingPipeline
//###############################################################
//***************************************************************
//1. GAM (Generalized Additive Models)
//***************************************************************
var trainer_gam = mlContext.Regression.Trainers
.Gam(labelColumnName: "LabelNormalized",
featureColumnName: "Features",
learningRate: 0.02,
numberOfIterations: 2100);
//Se añade el Algoritmo al pipeline de transformación de datos
IEstimator <ITransformer> trainingPipeline_gam = pipeline.Append(trainer_gam);
//***************************************************************
//2. GBA (Gradient Boosting Algorithm)
//***************************************************************
var trainer_boost = mlContext.Regression.Trainers
.FastTree(labelColumnName: "LabelNormalized",
featureColumnName: "Features",
numberOfLeaves: 20,
numberOfTrees: 100,
minimumExampleCountPerLeaf: 10,
learningRate: 0.2);
//Se añade el Algoritmo al pipeline de transformación de datos
IEstimator <ITransformer> trainingPipeline_boost = pipeline.Append(trainer_boost);
//###############################################################
//ENTRENAMIENTO DE LOS MODELOS
//###############################################################
Console.WriteLine($"\n************************************************************");
Console.WriteLine($"* Entrenamiento del Modelo calculado con el Algoritmo GAM ");
Console.WriteLine($"*-----------------------------------------------------------");
var watch_gam = System.Diagnostics.Stopwatch.StartNew();
var model_gam = trainingPipeline_gam.Fit(trainingDataView);
watch_gam.Stop();
var elapseds_gam = watch_gam.ElapsedMilliseconds * 0.001;
Console.WriteLine($"El entrenamiento GAM ha tardado: {elapseds_gam:#.##} s\n");
Console.WriteLine($"\n************************************************************");
Console.WriteLine($"* Entrenamiento del Modelo calculado con el Algoritmo GBA ");
Console.WriteLine($"*-----------------------------------------------------------");
var watch_boost = System.Diagnostics.Stopwatch.StartNew();
var model_boost = trainingPipeline_boost.Fit(trainingDataView);
watch_boost.Stop();
var elapseds_boost = watch_boost.ElapsedMilliseconds * 0.001;
Console.WriteLine($"El entrenamiento GBA ha tardado: {elapseds_boost:#.##} s\n");
//###############################################################
//EVALUACIÓN DE LOS MODELOS
//###############################################################
//Transformación del IDataView testDataView a paritr de ambos modelos
var predictions_gam = model_gam.Transform(testDataView);
var predictions_boost = model_boost.Transform(testDataView);
//Calculo de las métricas de cada Modelo
var metrics_gam = mlContext.Regression
.Evaluate(data: predictions_gam, labelColumnName: "LabelNormalized", scoreColumnName: "Score");
var metrics_boost = mlContext.Regression
.Evaluate(data: predictions_boost, labelColumnName: "LabelNormalized", scoreColumnName: "Score");
//Muestra las métricas GAM
Console.WriteLine($"\n************************************************************");
Console.WriteLine($"* Métricas para el Modelo calculado con el Algoritmo GAM ");
Console.WriteLine($"*-----------------------------------------------------------");
Console.WriteLine($"* GAM RSquared Score: {metrics_gam.RSquared:0.##}");
Console.WriteLine($"* GAM Root Mean Squared Error Score: {metrics_gam.RootMeanSquaredError:#.##}");
Console.WriteLine($"* GAM MAE Score: {metrics_gam.MeanAbsoluteError:#.##}");
Console.WriteLine($"* GAM MSE Score: {metrics_gam.MeanSquaredError:#.##}\n");
//Muestra las métricas GBA
Console.WriteLine($"\n************************************************************");
Console.WriteLine($"* Métricas para el Modelo calculado con el Algoritmo GBA ");
Console.WriteLine($"*-----------------------------------------------------------");
Console.WriteLine($"* GBA RSquared Score: {metrics_boost.RSquared:0.##}");
Console.WriteLine($"* GBA Root Mean Squared Error Score: {metrics_boost.RootMeanSquaredError:#.##}");
Console.WriteLine($"* GBA MAE Score: {metrics_boost.MeanAbsoluteError:#.##}");
Console.WriteLine($"* GBA MSE Score: {metrics_boost.MeanSquaredError:#.##}\n");
//###############################################################
//SELECCIÓN DEL MEJOR MODELO
//###############################################################
//Guardamos el Modelo para su posterior consumo
mlContext.Model.Save(model_boost, trainingDataView.Schema, _salida_modelPath);
}
//------------------------------------------------------------------------------------------------------------------
// Основная функция по построению регрессионных моделей для каждого из массивов и по получению прогнозных значений
public static void BuildPredictModel(ref float[] arrPredict1, ref float[] arrPredict2, ref float[] arrPredict3, ref float[] arrPredict4)
{
//Работа с первой моделью(первым массивом данных)
//Подготовка данных для модели
IDataView trainingDataView_1 = mlContext1.Data.LoadFromTextFile <ModelInput>(
path: TRAIN_DATA_FILEPATH_1,
hasHeader: true,
separatorChar: ',',
allowQuoting: true,
allowSparse: false);
IEstimator <ITransformer> dataProcessPipeline = mlContext1.Transforms.Concatenate("Features", new[] { "id" })
.Append(mlContext1.Transforms.NormalizeMinMax("Features", "Features"))
.AppendCacheCheckpoint(mlContext1);
// Выбор тренировочного алгоритма
var trainer = mlContext1.Regression.Trainers.LbfgsPoissonRegression(new LbfgsPoissonRegressionTrainer.Options()
{
DenseOptimizer = true, LabelColumnName = "name", FeatureColumnName = "Features"
});
IEstimator <ITransformer> trainingPipeline = dataProcessPipeline.Append(trainer);
// Тренировка модели
ITransformer model1 = trainingPipeline.Fit(trainingDataView_1);
// Оценка модели и вывод результатов оценки (для использования только в консоли)
//var crossValidationResults = mlContext1.Regression.CrossValidate(trainingDataView_1, trainingPipeline, numberOfFolds: 5, labelColumnName: "name");
//PrintStatistics(crossValidationResults);
// Сохранение модели
mlContext1.Model.Save(model1, trainingDataView_1.Schema, GetAbsolutePath(MODEL_FILEPATH_1));
// Получение множественного прогноза (4 прогнозных значения)
IDataView inputPredict = mlContext1.Data.LoadFromEnumerable <ModelInput>(inputModelData);
IDataView predictions = model1.Transform(inputPredict);
arrPredict1 = predictions.GetColumn <float>("Score").ToArray();
//------------------------------------------------------------------------------------------------------------------
// Работа со второй моделью (вторым массивом данных)
// Подготовка данных для модели
IDataView trainingDataView_2 = mlContext2.Data.LoadFromTextFile <ModelInput>(
path: TRAIN_DATA_FILEPATH_2,
hasHeader: true,
separatorChar: ',',
allowQuoting: true,
allowSparse: false);
IEstimator <ITransformer> dataProcessPipeline2 = mlContext2.Transforms.Concatenate("Features", new[] { "id" })
.Append(mlContext2.Transforms.NormalizeMinMax("Features", "Features"))
.AppendCacheCheckpoint(mlContext2);
// Выбор тренировочного алгоритма
var trainer2 = mlContext2.Regression.Trainers.LbfgsPoissonRegression(new LbfgsPoissonRegressionTrainer.Options()
{
DenseOptimizer = true, LabelColumnName = "name", FeatureColumnName = "Features"
});
IEstimator <ITransformer> trainingPipeline2 = dataProcessPipeline2.Append(trainer2);
// Тренировка модели
ITransformer model2 = trainingPipeline2.Fit(trainingDataView_2);
// Оценка модели и вывод результатов оценки (для использования только в консоли)
//var crossValidationResults2 = mlContext2.Regression.CrossValidate(trainingDataView_2, trainingPipeline2, numberOfFolds: 5, labelColumnName: "name");
//PrintStatistics(crossValidationResults2);
// Сохранение модели
mlContext2.Model.Save(model2, trainingDataView_2.Schema, GetAbsolutePath(MODEL_FILEPATH_2));
// Получение множественного прогноза (4 прогнозных значения)
IDataView inputPredict2 = mlContext2.Data.LoadFromEnumerable <ModelInput>(inputModelData);
IDataView predictions2 = model2.Transform(inputPredict2);
arrPredict2 = predictions2.GetColumn <float>("Score").ToArray();
//------------------------------------------------------------------------------------------------------------------
// Работа с третьей моделью (третьим массивом данных)
// Подготовка данных для модели
IDataView trainingDataView_3 = mlContext3.Data.LoadFromTextFile <ModelInput>(
path: TRAIN_DATA_FILEPATH_3,
hasHeader: true,
separatorChar: ',',
allowQuoting: true,
allowSparse: false);
IEstimator <ITransformer> dataProcessPipeline3 = mlContext3.Transforms.Concatenate("Features", new[] { "id" })
.Append(mlContext3.Transforms.NormalizeMinMax("Features", "Features"))
.AppendCacheCheckpoint(mlContext3);
// Выбор тренировочного алгоритма
var trainer3 = mlContext3.Regression.Trainers.LbfgsPoissonRegression(new LbfgsPoissonRegressionTrainer.Options()
{
DenseOptimizer = true, LabelColumnName = "name", FeatureColumnName = "Features"
});
IEstimator <ITransformer> trainingPipeline3 = dataProcessPipeline3.Append(trainer3);
// Тренировка модели
ITransformer model3 = trainingPipeline3.Fit(trainingDataView_3);
// Оценка модели и вывод результатов оценки (для использования только в консоли)
//var crossValidationResults3 = mlContext3.Regression.CrossValidate(trainingDataView_3, trainingPipeline3, numberOfFolds: 5, labelColumnName: "name");
//PrintStatistics(crossValidationResults3);
// Сохранение модели
mlContext3.Model.Save(model3, trainingDataView_3.Schema, GetAbsolutePath(MODEL_FILEPATH_3));
// Получение множественного прогноза (4 прогнозных значения)
IDataView inputPredict3 = mlContext3.Data.LoadFromEnumerable <ModelInput>(inputModelData);
IDataView predictions3 = model3.Transform(inputPredict3);
arrPredict3 = predictions3.GetColumn <float>("Score").ToArray();
//------------------------------------------------------------------------------------------------------------------
// Работа с четвертой моделью (четвертым массивом данных)
// Подготовка данных для модели
IDataView trainingDataView_4 = mlContext4.Data.LoadFromTextFile <ModelInput>(
path: TRAIN_DATA_FILEPATH_4,
hasHeader: true,
separatorChar: ',',
allowQuoting: true,
allowSparse: false);
IEstimator <ITransformer> dataProcessPipeline4 = mlContext4.Transforms.Concatenate("Features", new[] { "id" })
.Append(mlContext4.Transforms.NormalizeMinMax("Features", "Features"))
.AppendCacheCheckpoint(mlContext4);
// Выбор тренировочного алгоритма
var trainer4 = mlContext4.Regression.Trainers.LbfgsPoissonRegression(new LbfgsPoissonRegressionTrainer.Options()
{
DenseOptimizer = true, LabelColumnName = "name", FeatureColumnName = "Features"
});
IEstimator <ITransformer> trainingPipeline4 = dataProcessPipeline4.Append(trainer4);
// Тренировка модели
ITransformer model4 = trainingPipeline4.Fit(trainingDataView_4);
// Оценка модели и вывод результатов оценки (для использования только в консоли)
//var crossValidationResults4 = mlContext4.Regression.CrossValidate(trainingDataView_4, trainingPipeline4, numberOfFolds: 5, labelColumnName: "name");
//PrintStatistics(crossValidationResults4);
// Сохранение модели
mlContext4.Model.Save(model4, trainingDataView_4.Schema, GetAbsolutePath(MODEL_FILEPATH_4));
// Получение множественного прогноза (4 прогнозных значения)
IDataView inputPredict4 = mlContext4.Data.LoadFromEnumerable <ModelInput>(inputModelData);
IDataView predictions4 = model4.Transform(inputPredict4);
arrPredict4 = predictions4.GetColumn <float>("Score").ToArray();
}
private static ITransformer Train(IDataView trainingDataView, IEstimator <ITransformer> pipeLine)
{
// Train your model based on the data set
return(pipeLine.Fit(trainingDataView));
}
public ITransformer Train(IDataView trainingData)
{
TrainedModel = _trainingPipeline.Fit(trainingData);
return(TrainedModel);
}
/// <summary>
/// Train the model.
/// </summary>
/// <param name="trainingPipeline">Pipeline for model training.</param>
/// <param name="trainData">Training data view.</param>
/// <returns>The trained model.</returns>
public static ITransformer TrainModel(IEstimator <ITransformer> trainingPipeline, IDataView trainData)
{
return(trainingPipeline.Fit(trainData));
}
//creates the training algorithm class, trains the model, predicts area from training data then returns model
public static void TrainAndSaveModel(IDataView trainingDataView, IEstimator <ITransformer> pipeline)
{
var trainedModel = pipeline.Fit(trainingDataView);
_mlContext.Model.Save(trainedModel, trainingDataView.Schema, _modelPath);
}
static void Main(string[] args)
{
var mlContext = new MLContext();
IDataView attritionData = mlContext.Data.LoadFromTextFile <Employee>(path: "./data/attrition.csv", hasHeader: true, separatorChar: ',');
var split = mlContext.Data.TrainTestSplit(attritionData, testFraction: 0.2);
var trainData = split.TrainSet;
var testData = split.TestSet;
var numFields = attritionData.Schema.AsEnumerable()
.Select(column => new { column.Name, column.Type })
.Where(column => (column.Name != nameof(Employee.Attrition)) && (column.Type.ToString() == "Single"))
.ToArray();
var numFieldNames = numFields.AsEnumerable()
.Select(column => column.Name)
.ToList();
var oheFieldNames = new List <string>();
oheFieldNames.Add("OHE-" + nameof(Employee.BusinessTravel));
oheFieldNames.Add("OHE-" + nameof(Employee.Department));
oheFieldNames.Add("OHE-" + nameof(Employee.EducationField));
oheFieldNames.Add("OHE-" + nameof(Employee.MaritalStatus));
oheFieldNames.Add("OHE-" + nameof(Employee.JobLevel));
oheFieldNames.Add("OHE-" + nameof(Employee.JobRole));
oheFieldNames.Add("OHE-" + nameof(Employee.OverTime));
var allFeatureFields = new List <string>();
allFeatureFields.AddRange(oheFieldNames);
string[] numFeatures = numFieldNames.ToArray();
allFeatureFields.AddRange(numFeatures);
string[] allFeatureNames = allFeatureFields.ToArray();
IEstimator <ITransformer> featurizePipeline = mlContext.Transforms.Categorical.OneHotEncoding(
new[]
{
new InputOutputColumnPair("OHE-" + nameof(Employee.BusinessTravel), nameof(Employee.BusinessTravel)),
new InputOutputColumnPair("OHE-" + nameof(Employee.Department), nameof(Employee.Department)),
new InputOutputColumnPair("OHE-" + nameof(Employee.EducationField), nameof(Employee.EducationField)),
new InputOutputColumnPair("OHE-" + nameof(Employee.MaritalStatus), nameof(Employee.MaritalStatus)),
new InputOutputColumnPair("OHE-" + nameof(Employee.JobLevel), nameof(Employee.JobLevel)),
new InputOutputColumnPair("OHE-" + nameof(Employee.JobRole), nameof(Employee.JobRole)),
new InputOutputColumnPair("OHE-" + nameof(Employee.OverTime), nameof(Employee.OverTime))
}, OneHotEncodingEstimator.OutputKind.Indicator);
featurizePipeline = featurizePipeline.Append(mlContext.Transforms.Concatenate("Features", allFeatureNames))
.Append(mlContext.Transforms.NormalizeMinMax("Features", "Features"));
ConsoleHelper.ConsoleWriteHeader("=============== Begin to train the model ===============");
var trainer = mlContext.BinaryClassification.Trainers.SdcaLogisticRegression(
labelColumnName: nameof(Employee.Attrition),
featureColumnName: "Features");
/* ----- Tried with other trainers below and compared the outcome ------ */
// var trainer = mlContext.BinaryClassification.Trainers.LightGbm(labelColumnName: nameof(Employee.Attrition), featureColumnName: "Features");
// var trainer = mlContext.BinaryClassification.Trainers.FastTree(labelColumnName: nameof(Employee.Attrition), featureColumnName: "Features");
// var trainer = mlContext.BinaryClassification.Trainers.LbfgsLogisticRegression(labelColumnName: nameof(Employee.Attrition), featureColumnName: "Features");
// var trainer = mlContext.BinaryClassification.Trainers.SgdCalibrated(labelColumnName: nameof(Employee.Attrition), featureColumnName: "Features");
/* ------------------------------------------------------------------- */
var trainPipeline = featurizePipeline.Append(trainer);
var trainedModel = trainPipeline.Fit(trainData);
ConsoleHelper.ConsoleWriteHeader("=============== Trained model successfully ===============");
/*
* var viewTrainPipeline = mlContext.Transforms
* .CalculateFeatureContribution(model.LastTransformer)
* .Fit(dataPipeline.Fit(trainData).Transform(trainData));
*/
Console.WriteLine("===== Evaluating Model's accuracy with Test data =====");
var testDataPredictions = trainedModel.Transform(testData);
var evaluateMetrics = mlContext.BinaryClassification.Evaluate(data: testDataPredictions,
labelColumnName: nameof(Employee.Attrition),
scoreColumnName: "Score");
ConsoleHelper.PrintBinaryClassificationMetrics(trainedModel.ToString(), evaluateMetrics);
Console.WriteLine("===== Permutation Test =====");
var permuteTestData = featurizePipeline.Fit(trainData).Transform(trainData);
var permutationMetrics = mlContext.BinaryClassification.PermutationFeatureImportance(
predictionTransformer: trainedModel.LastTransformer,
data: permuteTestData,
labelColumnName: nameof(Employee.Attrition),
permutationCount: 50);
var mapFields = new List <string>();
for (int i = 0; i < allFeatureNames.Count(); i++)
{
var slotField = new VBuffer <ReadOnlyMemory <char> >();
if (permuteTestData.Schema[allFeatureNames[i]].HasSlotNames())
{
permuteTestData.Schema[allFeatureNames[i]].GetSlotNames(ref slotField);
for (int j = 0; j < slotField.Length; j++)
{
mapFields.Add(allFeatureNames[i]);
}
}
else
{
mapFields.Add(allFeatureNames[i]);
}
}
// Now let's look at which features are most important to the model
// overall. Get the feature indices sorted by their impact on AUC.
var sortedIndices = permutationMetrics
.Select((metrics, index) => new { index, metrics.AreaUnderRocCurve })
.OrderByDescending(
feature => Math.Abs(feature.AreaUnderRocCurve.Mean));
foreach (var feature in sortedIndices)
{
Console.WriteLine($"{mapFields[feature.index],-20}|\t{Math.Abs(feature.AreaUnderRocCurve.Mean):F6}");
}
}
// </SnippetDeclareGlobalVariables>
static void Main(string[] args)
{
// Create MLContext to be shared across the model creation workflow objects
// <SnippetCreateMLContext>
MLContext mlContext = new MLContext();
// </SnippetCreateMLContext>
// Dictionary to encode words as integers.
// <SnippetCreateLookupMap>
var lookupMap = mlContext.Data.LoadFromTextFile(Path.Combine(_modelPath, "imdb_word_index.csv"),
columns: new[]
{
new TextLoader.Column("Words", DataKind.String, 0),
new TextLoader.Column("Ids", DataKind.Int32, 1),
},
separatorChar: ','
);
// </SnippetCreateLookupMap>
// The model expects the input feature vector to be a fixed length vector.
// This action resizes the variable length array generated by the lookup map
// to a fixed length vector. If there are less than 600 words in the sentence,
// the remaining indices will be filled with zeros. If there are more than
// 600 words in the sentence, then the array is truncated at 600.
// <SnippetResizeFeatures>
Action <VariableLength, FixedLength> ResizeFeaturesAction = (s, f) =>
{
var features = s.VariableLengthFeatures;
Array.Resize(ref features, FeatureLength);
f.Features = features;
};
// </SnippetResizeFeatures>
// Load the TensorFlow model.
// <SnippetLoadTensorFlowModel>
TensorFlowModel tensorFlowModel = mlContext.Model.LoadTensorFlowModel(_modelPath);
// </SnippetLoadTensorFlowModel>
// <SnippetGetModelSchema>
DataViewSchema schema = tensorFlowModel.GetModelSchema();
Console.WriteLine(" =============== TensorFlow Model Schema =============== ");
var featuresType = (VectorDataViewType)schema["Features"].Type;
Console.WriteLine($"Name: Features, Type: {featuresType.ItemType.RawType}, Size: ({featuresType.Dimensions[0]})");
var predictionType = (VectorDataViewType)schema["Prediction/Softmax"].Type;
Console.WriteLine($"Name: Prediction/Softmax, Type: {predictionType.ItemType.RawType}, Size: ({predictionType.Dimensions[0]})");
// </SnippetGetModelSchema>
// <SnippetTokenizeIntoWords>
IEstimator <ITransformer> pipeline =
// Split the text into individual words
mlContext.Transforms.Text.TokenizeIntoWords("TokenizedWords", "ReviewText")
// </SnippetTokenizeIntoWords>
// <SnippetMapValue>
// Map each word to an integer value. The array of integer makes up the input features.
.Append(mlContext.Transforms.Conversion.MapValue("VariableLengthFeatures", lookupMap,
lookupMap.Schema["Words"], lookupMap.Schema["Ids"], "TokenizedWords"))
// </SnippetMapValue>
// <SnippetCustomMapping>
// Resize variable length vector to fixed length vector.
.Append(mlContext.Transforms.CustomMapping(ResizeFeaturesAction, "Resize"))
// </SnippetCustomMapping>
// <SnippetScoreTensorFlowModel>
// Passes the data to TensorFlow for scoring
.Append(tensorFlowModel.ScoreTensorFlowModel("Prediction/Softmax", "Features"))
// </SnippetScoreTensorFlowModel>
// <SnippetCopyColumns>
// Retrieves the 'Prediction' from TensorFlow and copies to a column
.Append(mlContext.Transforms.CopyColumns("Prediction", "Prediction/Softmax"));
// </SnippetCopyColumns>
// <SnippetCreateModel>
// Create an executable model from the estimator pipeline
IDataView dataView = mlContext.Data.LoadFromEnumerable(new List <MovieReview>());
ITransformer model = pipeline.Fit(dataView);
// </SnippetCreateModel>
// <SnippetCallPredictSentiment>
PredictSentiment(mlContext, model);
// </SnippetCallPredictSentiment>
}
// ===========================================================================================================
public void BuildTrainingPipelineAndModel(ClassificationMode classificationMode)
{
if (ErrorHasOccured)
{
return;
}
try
{
switch (classificationMode)
{
case ClassificationMode.OneVersusAll: // first time this project was builot
// Data process configuration with pipeline data transformations
EstimatorChain <NormalizingTransformer> dataProcessPipeline1 =
_mlContext.Transforms.Conversion.MapValueToKey("Label", "Label")
.Append(_mlContext.Transforms.Concatenate(
outputColumnName: "Features",
inputColumnNames: FeatureNames.ToArray()))
.Append(_mlContext.Transforms.NormalizeMinMax("Features", "Features"))
.AppendCacheCheckpoint(_mlContext);
// Set the training algorithm
EstimatorChain <KeyToValueMappingTransformer> trainer1 =
_mlContext.MulticlassClassification.Trainers.OneVersusAll(
_mlContext.BinaryClassification.Trainers.AveragedPerceptron(labelColumnName: "Label",
numberOfIterations: 10, featureColumnName: "Features"), labelColumnName: "Label")
.Append(_mlContext.Transforms.Conversion.MapKeyToValue("PredictedLabel", "PredictedLabel"));
IEstimator <ITransformer> trainingPipeline1 =
dataProcessPipeline1.Append(trainer1);
// Train Model
_mlModel = trainingPipeline1.Fit(_trainingDataView);
break;
case ClassificationMode.LightGbm:
// Data process configuration with pipeline data transformations
EstimatorChain <ColumnConcatenatingTransformer> dataProcessPipeline2 =
_mlContext.Transforms.Conversion.MapValueToKey("Label", "Label")
.Append(_mlContext.Transforms.Concatenate(
outputColumnName: "Features",
inputColumnNames: FeatureNames.ToArray()));
// Set the training algorithm
EstimatorChain <KeyToValueMappingTransformer> trainer2 =
_mlContext.MulticlassClassification.Trainers.LightGbm(labelColumnName: "Label", featureColumnName: "Features")
.Append(_mlContext.Transforms.Conversion.MapKeyToValue("PredictedLabel", "PredictedLabel"));
EstimatorChain <TransformerChain <KeyToValueMappingTransformer> > trainingPipeline2 =
dataProcessPipeline2.Append(trainer2);
// Train Model
_mlModel = trainingPipeline2.Fit(_trainingDataView);
break;
default:
throw new ArgumentOutOfRangeException(nameof(classificationMode), classificationMode, null);
}
// TODO
//// Cross-Validate with single dataset (since we don't have two datasets, one for training and for evaluate)
//// in order to evaluate and get the model's accuracy metrics
//Console.WriteLine("=============== Cross-validating to get model's accuracy metrics ===============");
//var crossValidationResults = mlContext.MulticlassClassification.CrossValidate(trainingDataView, trainingPipeline, numberOfFolds: 5, labelColumnName: "Label");
//PrintMulticlassClassificationFoldsAverageMetrics(crossValidationResults);
}
catch (Exception ex)
{
Debug.WriteLine(ex.Message);
ErrorHasOccured = true;
FailureInformation = ex.Message;
return;
}
}
public static IDataView FitAndTransform(this IEstimator <ITransformer> est, IDataView data) => est.Fit(data).Transform(data);
static ITransformer TrainModel(IDataView trainingData, IEstimator <ITransformer> trainingPipeline)
{
return(trainingPipeline.Fit(trainingData));
}
private void MixMatch(string dataPath)
{
// Create a new context for ML.NET operations. It can be used for exception tracking and logging,
// as a catalog of available operations and as the source of randomness.
var mlContext = new MLContext();
// Read the data as an IDataView.
// First, we define the loader: specify the data columns and where to find them in the text file.
var loader = mlContext.Data.CreateTextLoader(ctx => (
// The four features of the Iris dataset.
SepalLength: ctx.LoadFloat(0),
SepalWidth: ctx.LoadFloat(1),
PetalLength: ctx.LoadFloat(2),
PetalWidth: ctx.LoadFloat(3),
// Label: kind of iris.
Label: ctx.LoadText(4)
),
// Default separator is tab, but the dataset has comma.
separator: ',');
// Read the data.
var data = loader.Load(dataPath);
// Build the pre-processing pipeline.
var pipeline = loader.MakeNewEstimator()
.Append(r => (
// Convert string label to a key.
Label: r.Label.ToKey(),
// Concatenate all the features together into one column 'Features'.
Features: r.SepalLength.ConcatWith(r.SepalWidth, r.PetalLength, r.PetalWidth)));
// Now, at the time of writing, there is no static pipeline for OVA (one-versus-all). So, let's
// append the OVA learner to the dynamic pipeline.
IEstimator <ITransformer> dynamicPipe = pipeline.AsDynamic;
// Create a binary classification trainer.
var binaryTrainer = mlContext.BinaryClassification.Trainers.AveragedPerceptron("Label", "Features");
// Append the OVA learner to the pipeline.
dynamicPipe = dynamicPipe.Append(mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer));
// At this point, we have a choice. We could continue working with the dynamically-typed pipeline, and
// ultimately call dynamicPipe.Fit(data.AsDynamic) to get the model, or we could go back into the static world.
// Here's how we go back to the static pipeline:
var staticFinalPipe = dynamicPipe.AssertStatic(mlContext,
// Declare the shape of the input. As you can see, it's identical to the shape of the loader:
// four float features and a string label.
c => (
SepalLength: c.R4.Scalar,
SepalWidth: c.R4.Scalar,
PetalLength: c.R4.Scalar,
PetalWidth: c.R4.Scalar,
Label: c.Text.Scalar),
// Declare the shape of the output (or a relevant subset of it).
// In our case, we care only about the predicted label column (a key type), and scores (vector of floats).
c => (
Score: c.R4.Vector,
// Predicted label is a key backed by uint, with text values (since original labels are text).
PredictedLabel: c.KeyU4.TextValues.Scalar))
// Convert the predicted label from key back to the original string value.
.Append(r => r.PredictedLabel.ToValue());
// Train the model in a statically typed way.
var model = staticFinalPipe.Fit(data);
// And here is how we could've stayed in the dynamic pipeline and train that way.
dynamicPipe = dynamicPipe.Append(new KeyToValueMappingEstimator(mlContext, "PredictedLabel"));
var dynamicModel = dynamicPipe.Fit(data.AsDynamic);
// Now 'dynamicModel', and 'model.AsDynamic' are equivalent.
}
public void Train(string trainingFileName, string testFileName)
{
System.Diagnostics.Debug.WriteLine("Reached Train Method");
//Check if training file exists
if (!File.Exists(trainingFileName))
{
System.Diagnostics.Debug.WriteLine($"Failed to find training data file ({trainingFileName}");
return;
}
//Check if test file exists
if (!File.Exists(testFileName))
{
System.Diagnostics.Debug.WriteLine($"Failed to find test data file ({testFileName}");
return;
}
//Convert training file into IDataView object (ready for processing)
var trainingDataView = MlContext.Data.LoadFromTextFile <CarInventory>(trainingFileName, ',', hasHeader: false);
//Normalise Mean Variance on the inputted values
IEstimator <ITransformer> dataProcessPipeline = MlContext.Transforms
.Concatenate("Features", typeof(CarInventory)
.ToPropertyList <CarInventory>(nameof(CarInventory.Label)))
.Append(MlContext.Transforms.NormalizeMeanVariance(inputColumnName: "Features", outputColumnName: "FeaturesNormalizedByMeanVar"));
//Create a trainer object with the label from the car inventory class + normalised mean variance
var trainer = MlContext.BinaryClassification.Trainers.FastTree(
labelColumnName: nameof(CarInventory.Label),
featureColumnName: "FeaturesNormalizedByMeanVar",
numberOfLeaves: 2,
numberOfTrees: 800,
minimumExampleCountPerLeaf: 1,
learningRate: 0.2);
//Append the trainer to the pipeline
var trainingPipeline = dataProcessPipeline.Append(trainer);
//Save the model
var trainedModel = trainingPipeline.Fit(trainingDataView);
MlContext.Model.Save(trainedModel, trainingDataView.Schema, ModelPath);
//Evaluate the model like we trained it
var evaluationPipeline = trainedModel.Append(MlContext.Transforms
.CalculateFeatureContribution(trainedModel.LastTransformer)
.Fit(dataProcessPipeline.Fit(trainingDataView).Transform(trainingDataView)));
var testDataView = MlContext.Data.LoadFromTextFile <CarInventory>(testFileName, ',', hasHeader: false);
var testSetTransform = evaluationPipeline.Transform(testDataView);
var modelMetrics = MlContext.BinaryClassification.Evaluate(data: testSetTransform,
labelColumnName: nameof(CarInventory.Label),
scoreColumnName: "Score");
System.Diagnostics.Debug.WriteLine($"Accuracy: {modelMetrics.Accuracy:P2}");
System.Diagnostics.Debug.WriteLine($"Area Under Curve: {modelMetrics.AreaUnderRocCurve:P2}");
System.Diagnostics.Debug.WriteLine($"Area under Precision recall Curve: {modelMetrics.AreaUnderPrecisionRecallCurve:P2}");
System.Diagnostics.Debug.WriteLine($"F1Score: {modelMetrics.F1Score:P2}");
System.Diagnostics.Debug.WriteLine($"LogLoss: {modelMetrics.LogLoss:#.##}");
System.Diagnostics.Debug.WriteLine($"LogLossReduction: {modelMetrics.LogLossReduction:#.##}");
System.Diagnostics.Debug.WriteLine($"PositivePrecision: {modelMetrics.PositivePrecision:#.##}");
System.Diagnostics.Debug.WriteLine($"PositiveRecall: {modelMetrics.PositiveRecall:#.##}");
System.Diagnostics.Debug.WriteLine($"NegativePrecision: {modelMetrics.NegativePrecision:#.##}");
System.Diagnostics.Debug.WriteLine($"NegativeRecall: {modelMetrics.NegativeRecall:P2}");
}
