Neural Feature Search
This is an anonymous repository which contains tools and experimental data for Neural Feature Search: A Neural Architecture for Automated Feature Engineering.
To use Neural Feature Search, the following tools and environment are required:
- Python 3
- javabridge (>= 1.0.14)
- Oracle JDK 1.8+
- Python wrapper for the Java machine learning workbench Weka using the javabridge library.
- Weka (3.9.2)
- Scikit-learn 0.19.1
- TensorFlow 1.10.0
Installation
Detailed instructions and links to videos on installing the library are located here.
Please note that you need a built environment to compile some libraries from source.
Run
python Main_sequency.py --dataset file_name