Automated Tuning & Optimization of Models
ATOM aims at automating the data science process. Data scientists spend a lot of their time on model building and hyperparameter tuning which can be a tedious process and requires lots of experience. By using ATOM, they can focus more on other data science tasks like problem engineering and feature building. This automation will help many industries relying on machine learning in their core activities.
Dependencies:
- hyperopt
- numpy
- pandas
- scikitlearn
- xgboost
- keras
- pyqt4
Starting point:
MainWin.py