---
title: "MLObjectDetector.ModelParameters.ValidationData.split(strategy:)"
framework: createml
role: symbol
role_heading: Case
path: "createml/mlobjectdetector/modelparameters-swift.struct/validationdata/split(strategy:)"
---

# MLObjectDetector.ModelParameters.ValidationData.split(strategy:)

A validation dataset Create ML derives by randomly selecting a portion of the object detector’s training dataset using the split strategy.

## Declaration

```swift
case split(strategy: MLSplitStrategy)
```

## Parameters

- `strategy`: An doc://com.apple.createml/documentation/CreateML/MLSplitStrategy instance the enumeration case uses to select a portion of the object detector’s training dataset as its associated value.

## See Also

### Designating validation data

- [MLObjectDetector.ModelParameters.ValidationData.dataFrame(_:imageColumn:annotationColumn:)](createml/mlobjectdetector/modelparameters-swift.struct/validationdata/dataframe(_:imagecolumn:annotationcolumn:).md)
- [MLObjectDetector.ModelParameters.ValidationData.dataSource(_:)](createml/mlobjectdetector/modelparameters-swift.struct/validationdata/datasource(_:).md)
- [MLObjectDetector.ModelParameters.ValidationData.table(_:imageColumn:annotationColumn:)](createml/mlobjectdetector/modelparameters-swift.struct/validationdata/table(_:imagecolumn:annotationcolumn:).md)
- [MLObjectDetector.ModelParameters.ValidationData.none](createml/mlobjectdetector/modelparameters-swift.struct/validationdata/none.md)