7eaddd9a54
ML - устранение неточностей и багов ML - оптимизация производительности DecisionTreeRegressor.Fit, RandomForestRegressor.Fit ML - тесты
26 lines
507 B
ObjectPascal
26 lines
507 B
ObjectPascal
uses MLABC;
|
|
|
|
begin
|
|
var ds := Datasets.Iris;
|
|
var df := ds.Data;
|
|
|
|
var pipe :=
|
|
DataPipeline.Build(
|
|
TaskKind.tkClassification,
|
|
ds.Target,
|
|
ds.Features,
|
|
new StandardScaler, // Matrix transformer
|
|
new LogisticRegression // Model
|
|
);
|
|
|
|
var (trainDf, testDf) := df.TrainTestSplit(0.2, seed := 3);
|
|
|
|
pipe.Fit(trainDf);
|
|
|
|
var pred := pipe.Predict(testDf);
|
|
|
|
var y := pipe.GetEncodedLabels(testDf);
|
|
|
|
Println('Точность:', Metrics.Accuracy(y, pred):0:3);
|
|
end.
|