Skip to content

Latest commit

 

History

History
19 lines (11 loc) · 793 Bytes

File metadata and controls

19 lines (11 loc) · 793 Bytes

Interpretable Machine Learning - Exoplanet Identification

This is the third part of my undergraduate final project. The main goal was to find how neural networks classify an exoplanet.

Lime itself does not support time series data, so Lime For Time was used [1]. This library was slightly modified to support neural networks built with Keras.

The model used was the CNN with Global View input from [2].

References

[1] Jonas Hering, Emanuel Metzenthin & Alexander Zenner; Lime for Time ; https://github.com/emanuel-metzenthin/Lime-For-Time

[2] https://github.com/DiogorPinheiro/Exoplanet-Identification

[![Stars][stars-shield]][stars-url] [![Issues][issues-shield]][issues-url]