VIRAJ THAKKAR
MY PROJECTS
Machine Learning application in High Energy Physics
December 2018
My dedication to study machine learning began from the above-mentioned project at TIFR, Mumbai in 2017 when I was discovering various ways to classify and optimize signal and background particles. With this motivation during my summer vacations in 2018, I decided to study machine learning (ML) online and obtained a certificate.
Usually, in high energy physics, we mainly use rectangular cuts optimization to reduce the background and obtain maximum significance possible. It occurred to me that using supervised ML algorithms will lead to a better classification as compared to the traditional approach because such algorithms will classify signal and background using non-linear decision boundaries. I had to privilege to carry out this work for my master's thesis in NISER under the supervision of
Dr. Bedangadas Mohanty Find out more about this exciting project here!
Optimization of search variables in Leptoquark production at the CMS experiment of CERN
July 2017
I had worked on this project when I got selected for the VSRP-summer internship at the Tata Institute of Fundamental Research, Mumbai under the guidance of Dr. Gagan Mohanty. The project aims for the search of hypothetically proposed particles called "leptoquarks" that have the dual properties of both leptons and quarks. The task of the project was to optimize the value of various search variables in order to maximize the signal to background ratio for a leptoquark of a particular mass. The signal optimization was performed on Monte Carlo simulated data. A specific range of values for the search parameters was obtained to search for the leptoquark signal at the CMS detector at LHC, CERN. Click below to find out more!
Spin Alignment of Vector Mesons and Angular Momentum
April 2017
My sixth-semester project at the National Institute of Science Education and Research(NISER), India with Dr. Bedangadas Mohanty was the first major project which involved big data analysis and data mining. It involved deriving complex mathematical equations and application of Einstein’s special relativity to study the kinematics of the particles and to calculate useful physical quantities. Overwhelming at first, my job was to extract the signal particles from around 5 million events. Find out more about this work here!
Machine Learning applications in Dark Matter Search
May 2019
Do you know one of those things that you always wondered about in childhood? like dark matter, dark energy, black holes and all that crazy magical stuff that is out there in the universe! I finally got my hands on this exciting research topic of Dark Matter search in the universe, which would also be my last project at NISER (I am graduating in May 2019, times are really changing!). Thanks to Dr. Bedangadas Mohanty again, I had the opportunity to work with the dark matter search SuperCDMS collaboration at NISER. What is dark matter? How does it play an important role in shaping the universe? What does the SuperCDMS experiment do? How does machine learning come into the picture? Find out more about this work here!
Non-linear and Chaotic systems
April 2018
I worked on nonlinear systems when I had taken a course on Nonlinear physics and chaos, and while pursuing an experimental physics lab at NISER. Nonlinear and chaotic systems are one of the most intriguing things I have ever studied. Chaos theory has applications in many areas like physics, engineering, finance, economics, and computer science.
A chaotic system is like a capricious girlfriend, a small change in the initial conditions can lead to a drastic change in their behavior. Interested to know more? Check out the link below!