PHOTONAI offers easy access to established machine learning algorithms.
You can set all parameters of the imported class as usual: e.g. add gamma='auto' to the PipelineElement to set the support vector machine's gamma parameter to 'auto'.
In addition, you can specify each parameter as a hyperparameter and define a value range or value list to find the optimal value, such as 'kernel': ['linear', 'rbf'] .
To build a custom pipeline, have a look at PHOTONAIs pre-registered processing- and learning algorithms. You can access algorithms for all purposes from several open-source packages. In addition, PHOTONAI offers several utility classes as well, such as linear statistical feature selection or sample pairing algorithms.
In addition you can specify hyperparameters as well as their value range in order to be optimized by the hyperparameter optimization strategy. Currently, PHOTONAI offers Grid-Search, Random Search and two frameworks for bayesian optimization.
1 2 3 4 5 6
1 2 3 4
Keras Neural Net
1 2 3 4 5 6 7 8