Skip to content

Hyperparamteroptimization for RNNs and CNNs #912

@sibre28

Description

@sibre28

Is your feature request related to a problem?

We currently only support hyperparameter optimization for FNN Models. This results out of missing functionality to measure the Quality of TimeSeries Predictions and Image Predictions.

Desired solution

Add support for TimeSeries and Images as Input in fit_by_exhaustive_search in NeuralNetworkClassifier and NeuralNetworkRegressor.

  • For Time Series implement Block Validation and include the window_size in some way.
  • For Images differentiate between the different prediction methods via the InputConversion

Possible alternatives (optional)

No response

Screenshots (optional)

No response

Additional Context (optional)

No response

Metadata

Metadata

Assignees

Labels

releasedIncluded in a release

Type

No type

Projects

Status

✔️ Done

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions