Feature Engineering and Feature Selection

Feature engineering and feature selection are two of the most important processes for developing accurate models. Feature engineering involves wrangling of raw data into attributes that enhance the predictive power of the model. Feature selection approaches allow the machine to identify the features that are most important in the model; which enhance predictive power when generalizing a machine learning model to unseen data.

The original promise of deep learning led many to believe that the days of needing to handcraft features would be over when building neural networks. However, Data Scientists have discovered that deep learning has simplified the types of feature engineering required over traditional machine learning models, it has not eliminated the need for handcrafting some features.

When working with smaller data sets, machine learning models will outperform deep learning models and the techniques to optimize the feature engineering and feature selection processes are crucial to achieving the full potential of the model.

In this meetup, Ben Fowler will demonstrate best practice techniques and libraries for feature engineering and feature selection. Ben holds a Master of Science in Data Science from Southern Methodist University and is a Data Scientist for JM Family Enterprises.