Python Machine Learning Cookbook - Second Edition

How it works...

MLBox builds the whole pipeline with the following three steps:

1. Preprocessing: All of the operations related to this phase make use of the mlbox.preprocessing sub-package. In this phase, we proceed to the reading and cleaning of the input file and then to the removal of the drift variables.
2. Optimization: All of the operations related to this phase make use of the sub-package mlbox.mlbox.optimisation. In this phase, the whole pipeline is optimized. The hyperparametric optimization method that's adopted uses the hyperopt library. This library creates a highly-dimensional space for the parameters to be optimized and chooses the best combination of parameters that lowers the validation score.
3. Prediction: All of the operations ...