MALT: Machine Learning for Transients

MALT is a classification pipeline based on the paper “Classification of Multiwavelength Transients with Machine Learning” by Sooknunan et al. (2018). It is a framework which allows the user to classify time series data. The user is free to choose the interpolation technique, feature extraction method, and the machine learning classifier to use.

The default pipeline is shown below. It uses Gaussian processes to interpolate the data, a wavelet feature extraction method and a random forest classifier.

Contents:

• How to install MALT
• Example Notebooks
  • Lightcurve Example
  • Dataset Example
• The MALT API reference
  • The Lightcurve class
  • The Dataset class
  • MALT interpolator
  • MALT feature extraction
  • MALT machine learning