Study of semi-visible jet using machine learning method

In this project, the dark matter problem has been investigated. The Hidden Valley theory predicts the existence of the so-called semi-visible jet. Semi-visible jets are composed of some fraction of invisible dark matter particles mixed with visible standard model particles. The alignment of MET with the subleading jet gives a unique signature and offers a new way to look for dark matter. It has just been searched for at Large Hadron Collider at CERN. However the difficulty to find out such signature from SM jets suggest the used of machine learning method. The goal is to produce images of the semivisible jet and see if they are different from usual jets from the SM using machine learning. These images have been preprocessed in order to make the learning process more efficient and finally they have been used as input for the Convolution Neural Network.