---
title: predictionWindowSize
framework: createml
role: symbol
role_heading: Instance Property
path: createml/mlhandactionclassifier/modelparameters-swift.struct/predictionwindowsize
---

# predictionWindowSize

The number of video frames the model training session uses to train a hand action classifier.

## Declaration

```swift
var predictionWindowSize: Int
```

## Discussion

Discussion Set to 60 to capture hand actions with a 2 second duration from videos that have a frame rate of 30 frames per second.

## See Also

### Accessing hand action training parameters

- [maximumIterations](createml/mlhandactionclassifier/modelparameters-swift.struct/maximumiterations.md)
- [batchSize](createml/mlhandactionclassifier/modelparameters-swift.struct/batchsize.md)
- [targetFrameRate](createml/mlhandactionclassifier/modelparameters-swift.struct/targetframerate.md)
- [algorithm](createml/mlhandactionclassifier/modelparameters-swift.struct/algorithm.md)
- [augmentationOptions](createml/mlhandactionclassifier/modelparameters-swift.struct/augmentationoptions.md)
- [validation](createml/mlhandactionclassifier/modelparameters-swift.struct/validation.md)
