SUSY Analysis using ML Techniques

SUSY analysis using Machine Learning Techniques. The following ML techniques were used in the study:

1. XGBClassifier (from XGBoost)
2. Deep Neural Network
3. Variational Autoencoder (work still in progress)
4. (To Do) Autoencoder

These ML techniques were trained using low-level features, high-level features and combination of both low-level and high level features.

Dataset [2]:

1. Signal: The process $\chi^{\pm}~ \rightarrow~ W^{\pm} ~\chi^{o}$ with W boson decaying to lepton and neutrino.
2. Background: Pair of W boson decaying to lepton and neutrino.

Distribution of input features

ROC Curve

XGBClassifier for low, high and combination of low and high level features

DNN for low, high and combination of low and high level features

References:

1. Baldi, P., Sadowski, P. & Whiteson, D. Searching for exotic particles in high-energy physics with deep learning. Nat Commun 5, 4308 (2014). https://doi.org/10.1038/ncomms5308
2. Dataset used from http://archive.ics.uci.edu/ml/datasets/SUSY (Monte Carlo Simulation)