public void TrainSentiment()
{
var env = new MLContext(seed: 1);
// Pipeline
var arguments = new TextLoader.Arguments()
{
Column = new TextLoader.Column[]
{
new TextLoader.Column()
{
Name = "Label",
Source = new[] { new TextLoader.Range()
{
Min = 0, Max = 0
} },
Type = DataKind.Num
},
new TextLoader.Column()
{
Name = "SentimentText",
Source = new[] { new TextLoader.Range()
{
Min = 1, Max = 1
} },
Type = DataKind.Text
}
},
HasHeader = true,
AllowQuoting = false,
AllowSparse = false
};
var loader = env.Data.ReadFromTextFile(_sentimentDataPath, arguments);
var text = TextFeaturizingEstimator.Create(env,
new TextFeaturizingEstimator.Arguments()
{
Column = new TextFeaturizingEstimator.Column
{
Name = "WordEmbeddings",
Source = new[] { "SentimentText" }
},
OutputTokens = true,
KeepPunctuations = false,
UsePredefinedStopWordRemover = true,
VectorNormalizer = TextFeaturizingEstimator.TextNormKind.None,
CharFeatureExtractor = null,
WordFeatureExtractor = null,
}, loader);
var trans = WordEmbeddingsExtractingTransformer.Create(env,
new WordEmbeddingsExtractingTransformer.Arguments()
{
Column = new WordEmbeddingsExtractingTransformer.Column[1]
{
new WordEmbeddingsExtractingTransformer.Column
{
Name = "Features",
Source = "WordEmbeddings_TransformedText"
}
},
ModelKind = WordEmbeddingsExtractingTransformer.PretrainedModelKind.Sswe,
}, text);
// Train
var trainer = new SdcaMultiClassTrainer(env, "Label", "Features", maxIterations: 20);
var predicted = trainer.Fit(trans);
_consumer.Consume(predicted);
}
public void TrainAndPredictSentimentModelWithDirectionInstantiationTestWithWordEmbedding()
{
var dataPath = GetDataPath(SentimentDataPath);
var testDataPath = GetDataPath(SentimentTestPath);
using (var env = new ConsoleEnvironment(seed: 1, conc: 1))
{
// Pipeline
var loader = TextLoader.ReadFile(env,
new TextLoader.Arguments()
{
Separator = "tab",
HasHeader = true,
Column = new[]
{
new TextLoader.Column("Label", DataKind.Num, 0),
new TextLoader.Column("SentimentText", DataKind.Text, 1)
}
}, new MultiFileSource(dataPath));
var text = TextFeaturizingEstimator.Create(env, new TextFeaturizingEstimator.Arguments()
{
Column = new TextFeaturizingEstimator.Column
{
Name = "WordEmbeddings",
Source = new[] { "SentimentText" }
},
OutputTokens = true,
KeepPunctuations = false,
StopWordsRemover = new PredefinedStopWordsRemoverFactory(),
VectorNormalizer = TextFeaturizingEstimator.TextNormKind.None,
CharFeatureExtractor = null,
WordFeatureExtractor = null,
},
loader);
var trans = WordEmbeddingsExtractingTransformer.Create(env, new WordEmbeddingsExtractingTransformer.Arguments()
{
Column = new WordEmbeddingsExtractingTransformer.Column[1]
{
new WordEmbeddingsExtractingTransformer.Column
{
Name = "Features",
Source = "WordEmbeddings_TransformedText"
}
},
ModelKind = WordEmbeddingsExtractingTransformer.PretrainedModelKind.Sswe,
}, text);
// Train
var trainer = new FastTreeBinaryClassificationTrainer(env, DefaultColumnNames.Label, DefaultColumnNames.Features, numLeaves: 5, numTrees: 5, minDatapointsInLeaves: 2);
var trainRoles = new RoleMappedData(trans, label: "Label", feature: "Features");
var pred = trainer.Train(trainRoles);
// Get scorer and evaluate the predictions from test data
IDataScorerTransform testDataScorer = GetScorer(env, trans, pred, testDataPath);
var metrics = EvaluateBinary(env, testDataScorer);
// SSWE is a simple word embedding model + we train on a really small dataset, so metrics are not great.
Assert.Equal(.6667, metrics.Accuracy, 4);
Assert.Equal(.71, metrics.Auc, 1);
Assert.Equal(.58, metrics.Auprc, 2);
// Create prediction engine and test predictions
var model = env.CreateBatchPredictionEngine <SentimentData, SentimentPrediction>(testDataScorer);
var sentiments = GetTestData();
var predictions = model.Predict(sentiments, false);
Assert.Equal(2, predictions.Count());
Assert.True(predictions.ElementAt(0).Sentiment);
Assert.True(predictions.ElementAt(1).Sentiment);
// Get feature importance based on feature gain during training
var summary = ((FeatureWeightsCalibratedPredictor)pred).GetSummaryInKeyValuePairs(trainRoles.Schema);
Assert.Equal(1.0, (double)summary[0].Value, 1);
}
}
public void TrainSentiment()
{
using (var env = new ConsoleEnvironment(seed: 1))
{
// Pipeline
var loader = TextLoader.ReadFile(env,
new TextLoader.Arguments()
{
AllowQuoting = false,
AllowSparse = false,
Separator = "tab",
HasHeader = true,
Column = new[]
{
new TextLoader.Column()
{
Name = "Label",
Source = new [] { new TextLoader.Range()
{
Min = 0, Max = 0
} },
Type = DataKind.Num
},
new TextLoader.Column()
{
Name = "SentimentText",
Source = new [] { new TextLoader.Range()
{
Min = 1, Max = 1
} },
Type = DataKind.Text
}
}
}, new MultiFileSource(_sentimentDataPath));
var text = TextFeaturizingEstimator.Create(env,
new TextFeaturizingEstimator.Arguments()
{
Column = new TextFeaturizingEstimator.Column
{
Name = "WordEmbeddings",
Source = new[] { "SentimentText" }
},
OutputTokens = true,
KeepPunctuations = false,
StopWordsRemover = new PredefinedStopWordsRemoverFactory(),
VectorNormalizer = TextFeaturizingEstimator.TextNormKind.None,
CharFeatureExtractor = null,
WordFeatureExtractor = null,
}, loader);
var trans = WordEmbeddingsExtractingTransformer.Create(env,
new WordEmbeddingsExtractingTransformer.Arguments()
{
Column = new WordEmbeddingsExtractingTransformer.Column[1]
{
new WordEmbeddingsExtractingTransformer.Column
{
Name = "Features",
Source = "WordEmbeddings_TransformedText"
}
},
ModelKind = WordEmbeddingsExtractingTransformer.PretrainedModelKind.Sswe,
}, text);
// Train
var trainer = new SdcaMultiClassTrainer(env, "Label", "Features", maxIterations: 20);
var trainRoles = new RoleMappedData(trans, label: "Label", feature: "Features");
var predicted = trainer.Train(trainRoles);
_consumer.Consume(predicted);
}
}
