Gravitational Lens Finding for Dark Matter Substructure Pipeline

Machine learning techniques are regarded to have the potential to help researchers better comprehend dark matter. Convolutional Neural Networks (CNNs) have previously been used to successfully distinguish between distinct types of substructures such as no substructure, vortex substructure, and spherical substructure. In this task, we'll classify images based on whether or not they have a strong lensing effect. Strong Gravitational Lensing allows us to analyse the substructure of dark matter, allowing us to gain a deeper understanding of its underlying nature. The observed strong lens images may be utilized in the Deeplense pipeline for dark matter substructure classification, anomaly detection, and interpretation. Images may be processed using a variety of algorithms. We'll try out several types of models, fine-tune them to obtain good outcomes, and compare them. Because a single model may not be sufficient, we can combine the findings of several models using ensembling to obtain a more robust and better outcome. We'll document and record every outcome as we experiment with our models. We will also provide starter notebooks describing how to train and test the models so that it will be easy for anyone to understand our work.