---
title: "init(validation:maxIterations:augmentation:algorithm:)"
framework: createml
role: symbol
role_heading: Initializer
path: "createml/mlimageclassifier/modelparameters-swift.struct/init(validation:maxiterations:augmentation:algorithm:)"
---

# init(validation:maxIterations:augmentation:algorithm:)

Creates model training parameters.

## Declaration

```swift
init(validation: MLImageClassifier.ModelParameters.ValidationData = __Defaults.validation, maxIterations: Int = __Defaults.maximumIterations, augmentation: MLImageClassifier.ImageAugmentationOptions, algorithm: MLImageClassifier.ModelParameters.ModelAlgorithmType = __Defaults.algorithm)
```

## Parameters

- `validation`: Labeled data that the model evaluates on for validation. The default is .split(strategy: .automatic).
- `maxIterations`: The maximum number of training iterations to use during training. The default is 25.
- `augmentation`: The image augmentation options to use to increase the training data variety. If no data augmentation needs to be applied, use [] as input. Otherwise, inputs take the form [.crop, .blur].
- `algorithm`: The type of model algorithm to use for training. The default is a logistic regression classifier with a sceneprint(revision: 1) feature extractor.

## See Also

### Creating parameters

- [init(featureExtractor:validation:maxIterations:augmentationOptions:)](createml/mlimageclassifier/modelparameters-swift.struct/init(featureextractor:validation:maxiterations:augmentationoptions:).md)
- [init(featureExtractor:validationData:maxIterations:augmentationOptions:)](createml/mlimageclassifier/modelparameters-swift.struct/init(featureextractor:validationdata:maxiterations:augmentationoptions:).md)
- [MLImageClassifier.ModelParameters.ClassifierType](createml/mlimageclassifier/modelparameters-swift.struct/classifiertype.md)
- [MLImageClassifier.ModelParameters.ModelAlgorithmType](createml/mlimageclassifier/modelparameters-swift.struct/modelalgorithmtype.md)