A Modified Framework for Automated Feature Extraction using Deep Feature Synthesis | Original Article

ABSTRACT:

In recent years, advances in machine learning have led to various innovations in to automating various iterative and time-consuming tasks. One such tasks that data scientists carries out is feature extraction. Feature Extraction is time consuming and requires domain knowledge and chances are few features can be missed. As a result, automating this process will provide ease for data scientists as all possible features can be generated. Automated Feature Engineering majorly focuses on reducing the time require to generate features which can be used to train the models. As a result, a framework is provided to reduce the time required for feature extraction by considering only candidate set of features as an input that also helps to generate only useful features.