Home - SashaMalysheva/AutoML GitHub Wiki

Welcome to the AutoML wiki!

##AutoML AtoML provides out-of-the-box supervised machine learning. It's automatically searches for the right learning algorithm for a current machine learning dataset and optimizes its hyperparameters.

##Auto-sklearn Auto-sklearn is an open-source Python tool that automatically determines effective machine learning pipelines for classification and regression datasets.

Auto-sklearn includes 15 ML algorithms, 14 preprocessing methods, and all their respective hyperparameters, yielding a total of 110 hyperparameters; the resulting space is illustrated in Figure 1, where only the hyperparameters in grey boxes are active.

####What important Auto-sklearn comes with a database of previous optimization runs on 140 diverse datasets from OpenML. For a new dataset, it first identifies the most similar datasets and starts from the saved best settings for those.

Also It automatically construct ensembles: instead of returning a single hyperparameter setting, it construct ensembles from the models trained during the Bayesian optimization with using Ensemble Selection.

Diagram show importance of this improvement