private static void CreateCountryModel(MLContext mlContext, string dataPath, string outputModelPath)
{
ConsoleExtensions.ConsoleWriteHeader("Training country forecasting model");
var trainindDataView = mlContext.Data.LoadFromTextFile <CountryData>(path: dataPath, hasHeader: true, separatorChar: ',');
var trainer = mlContext.Regression.Trainers.FastTreeTweedie("Label", "Features", learningRate: 0.2);
var trainingPipeline = mlContext.Transforms.Concatenate(outputColumnName: "NumFeatures", nameof(CountryData.year),
nameof(CountryData.month), nameof(CountryData.max), nameof(CountryData.min),
nameof(CountryData.std), nameof(CountryData.count), nameof(CountryData.sales),
nameof(CountryData.med), nameof(CountryData.prev))
.Append(mlContext.Transforms.Categorical.OneHotEncoding(outputColumnName: "CatFeatures", inputColumnName: nameof(CountryData.country)))
.Append(mlContext.Transforms.Concatenate(outputColumnName: "Features", "NumFeatures", "CatFeatures"))
.Append(mlContext.Transforms.CopyColumns(outputColumnName: "Label", inputColumnName: nameof(CountryData.next)))
.Append(trainer);
// 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.Regression.CrossValidate(data: trainindDataView, estimator: trainingPipeline, numberOfFolds: 6, labelColumnName: "Label");
ConsoleExtensions.PrintRegressionFoldsAverageMetrics(trainer.ToString(), crossValidationResults);
// Create and Train the model
var model = trainingPipeline.Fit(trainindDataView);
//Save model
mlContext.Model.Save(model, trainindDataView.Schema, outputModelPath);
}
/// <summary>
/// Predict samples using saved model
/// </summary>
/// <param name="outputModelPath">Model file path</param>
/// <param name="mlContext"></param>
/// <returns></returns>
public static void TestPrediction(MLContext mlContext, string outputModelPath = "country_sales_ml_model.zip")
{
ConsoleExtensions.ConsoleWriteHeader("Testing Country Sales Forecast model");
ITransformer trainedModel;
using (var stream = File.OpenRead(outputModelPath))
{
trainedModel = mlContext.Model.Load(stream, out var modelInputSchema);
}
var predictionEngine = mlContext.Model.CreatePredictionEngine <CountryData, CountrySalesPrediction>(trainedModel);
Console.WriteLine("** Testing Country 1 **");
// Build sample data
var dataSample = new CountryData()
{
country = "United Kingdom",
month = 10,
year = 2017,
med = 309.945F,
max = 587.902F,
min = 135.640F,
std = 1063.932092F,
prev = 856548.78F,
count = 1724,
sales = 873612.9F,
};
var prediction = predictionEngine.Predict(dataSample);
Console.WriteLine($"Country: {dataSample.country}, month to predict: {dataSample.month + 1}, year: {dataSample.year} - Real value (US$): {Math.Pow(6.0084501F, 10)}, Predicted Forecast (US$): {Math.Pow(prediction.Score, 10)}");
dataSample = new CountryData()
{
country = "United Kingdom",
month = 11,
year = 2017,
med = 288.72F,
max = 501.488F,
min = 134.5360F,
std = 707.5642F,
prev = 873612.9F,
count = 2387,
sales = 1019647.67F,
};
prediction = predictionEngine.Predict(dataSample);
Console.WriteLine($"Country: {dataSample.country}, month to predict: {dataSample.month + 1}, year: {dataSample.year} - Predicted Forecast (US$): {Math.Pow(prediction.Score, 10)}");
Console.WriteLine(" ");
Console.WriteLine("** Testing Country 2 **");
dataSample = new CountryData()
{
country = "United States",
month = 10,
year = 2017,
med = 400.17F,
max = 573.63F,
min = 340.395F,
std = 340.3959F,
prev = 4264.94F,
count = 10,
sales = 5322.56F
};
prediction = predictionEngine.Predict(dataSample);
Console.WriteLine($"Country: {dataSample.country}, month to predict: {dataSample.month + 1}, year: {dataSample.year} - Real value (US$): {Math.Pow(3.805769F, 10)}, Predicted Forecast (US$): {Math.Pow(prediction.Score, 10)}");
dataSample = new CountryData()
{
country = "United States",
month = 11,
year = 2017,
med = 317.9F,
max = 1135.99F,
min = 249.44F,
std = 409.75528F,
prev = 5322.56F,
count = 11,
sales = 6393.96F,
};
prediction = predictionEngine.Predict(dataSample);
Console.WriteLine($"Country: {dataSample.country}, month to predict: {dataSample.month + 1}, year: {dataSample.year} - Predicted Forecast (US$): {Math.Pow(prediction.Score, 10)}");
}
