Hi, there! I'm Sanchi, a Computer Science grad. Welcome to my GSoC 2022 Blog! Starting with this one, I'm going to be putting out a series of blogs about my journey so far and yet to come in this process.
Selecting the Organisation: What is CERN-HSF?
CERN-HSF is an umbrella organization that brings high-energy physics-related projects in GSoC from the likes of CERN, Lawrence Berkeley National Laboratory, TUM, FermiLab, Imperial College of London, among many others. This is one org specifically that caught my interest, for it is not only one of the biggest and most well-known research organization, but also because it is the place www was invented and Higgs Boson was discovered. To me, it is the most amazing place that I could work for!
Selecting the Project
Being a CS student, one of the ways I could contribute to this organization was through Google Summer of Code. For this, I started by going over the project ideas listed on the official page link. As I was deeply passionate about Deep Learning and its latest applications, my search for a project was primarily in that domain. The projects that caught my attention were Geant4 and ROOT-TMVA.
My Application for Geant4
First I contacted the mentors for proposal submission and they replied back with a 30 min test which was a thorough evaluation of Machine learning concepts on a deep level. Unfortunately, I didn't get accepted for the project.
My Application for ROOT-TMVA: Batch Generator For Training Machine Learning Models
TMVA stands for Tools for Multi-Variate Analysis and it is the ROOT library that provides the interfaces and implementation of different machine learning techniques, including neural networks, decision trees, SVM, etc. I was following this project for quite some time and was very interested to contribute If I got the opportunity to do so. Outlining my experience and ideas for the project, I contacted the mentors.
I received a set of introductory exercises regarding building ROOT from source and familiarizing myself with the ROOT TMVA Deep Learning Code. Upon their completion, I was given a green signal to submit the proposal for the project and I also had to follow up by completing a set of the GSoC project-specific exercises where I had to implement an example tutorial in Python to read a ROOT TTree, train an ML model and implement a prototype generator.
I shared my ideas on the project implementation, additional features, and a timeline for the project goals completion in the proposal. I kept in touch with the mentors throughout the application period for their guidance with the drafts of the proposal and then went on to submit the final draft. The effort proved fruitful and I got through!
Stay tuned for more blogs along this journey. À bientôt!
- Sanchi