<!--
{
  "availability" : [
    "macOS: 10.15.0 -"
  ],
  "documentType" : "symbol",
  "framework" : "CreateML",
  "identifier" : "/documentation/CreateML/MLActivityClassifier/init(trainingData:featureColumns:labelColumn:recordingFileColumn:parameters:)",
  "metadataVersion" : "0.1.0",
  "role" : "Initializer",
  "symbol" : {
    "kind" : "Initializer",
    "modules" : [
      "Create ML"
    ],
    "preciseIdentifier" : "s:8CreateML20MLActivityClassifierV12trainingData14featureColumns11labelColumn013recordingFileJ010parametersA2C0F6SourceO_SaySSGSSSgAlC15ModelParametersVtKcfc::OverloadGroup"
  },
  "title" : "init(trainingData:featureColumns:labelColumn:recordingFileColumn:parameters:)"
}
-->

# init(trainingData:featureColumns:labelColumn:recordingFileColumn:parameters:)

Creates an activity classifier with a training dataset represented by a data source.

```
init(trainingData: MLActivityClassifier.DataSource, featureColumns: [String], labelColumn: String? = nil, recordingFileColumn: String? = nil, parameters: MLActivityClassifier.ModelParameters = ModelParameters(validationData: nil)) throws
```

## Parameters

`trainingData`

An [`MLActivityClassifier.DataSource`](/documentation/CreateML/MLActivityClassifier/DataSource) instance.

`featureColumns`

The names of the columns in an annotation file that contain sensor data.

`labelColumn`

The name of the column in an annotation file that contains the activity labels if
`trainingData` uses
[`MLActivityClassifier.DataSource.directoryWithDataAndAnnotation(at:annotationFileName:timeStampColumn:labelStartTimeColumn:labelEndTimeColumn:)`](/documentation/CreateML/MLActivityClassifier/DataSource/directoryWithDataAndAnnotation(at:annotationFileName:timeStampColumn:labelStartTimeColumn:labelEndTimeColumn:)).

    The initializer ignores this parameter if     `trainingData`     uses
[`MLActivityClassifier.DataSource.labeledDirectories(at:)`](/documentation/CreateML/MLActivityClassifier/DataSource/labeledDirectories(at:))    .

`recordingFileColumn`

The name of the column in an annotation file that contains the data filenames if
`trainingData` uses
[`MLActivityClassifier.DataSource.directoryWithDataAndAnnotation(at:annotationFileName:timeStampColumn:labelStartTimeColumn:labelEndTimeColumn:)`](/documentation/CreateML/MLActivityClassifier/DataSource/directoryWithDataAndAnnotation(at:annotationFileName:timeStampColumn:labelStartTimeColumn:labelEndTimeColumn:)).

    The initializer ignores this parameter if     `trainingData`     uses
[`MLActivityClassifier.DataSource.labeledDirectories(at:)`](/documentation/CreateML/MLActivityClassifier/DataSource/labeledDirectories(at:))    .

`parameters`

An [`MLActivityClassifier.ModelParameters`](/documentation/CreateML/MLActivityClassifier/ModelParameters-swift.struct) instance you use to configure the
model for the training session.

## Discussion

Use this initializer to create an activity classifier with an [`MLActivityClassifier.DataSource`](/documentation/CreateML/MLActivityClassifier/DataSource). To configure
the training process, initialize the activity classifier with an
[`MLActivityClassifier.ModelParameters`](/documentation/CreateML/MLActivityClassifier/ModelParameters-swift.struct) instance. For example, you can explicitly define the
validation dataset instead of allowing the model to choose a random selection of your training data.
Alternatively, set [`validationData`](/documentation/CreateML/MLActivityClassifier/ModelParameters-swift.struct/validationData) to `nil` to allow the
activity classifier to choose the validation data for you from among your training data. This lets you set
other parameters — like [`maximumIterations`](/documentation/CreateML/MLActivityClassifier/ModelParameters-swift.struct/maximumIterations) and
[`batchSize`](/documentation/CreateML/MLActivityClassifier/ModelParameters-swift.struct/batchSize) — to nondefault values.

---

Copyright &copy; 2026 Apple Inc. All rights reserved. | [Terms of Use](https://www.apple.com/legal/internet-services/terms/site.html) | [Privacy Policy](https://www.apple.com/privacy/privacy-policy)