Automated feature engineering in the python stack.
TPOT throws the SKLearn kitchen sink at feature engineering.
- Aggregated data.
- Nothing you couldn't build in your own SKLearn pipeline.
- Uses genetic search algorithm.
auto_ml creates new features using deep learning.
- Aggregated data.
- Penultimate layer of ten "best" features from DNN appended to data.
Features Tools takes the most innovative approach.
- Pre-aggregated data!
- Recursively applies SQL style aggregations.
- Binder Notebook.