Projects

Ongoing Research Projects

• Cross corelation between GL and SZ maps to constraint Cosmological parameters

  Supervisor: Subhabrata Majumdar

  TIFR, Mumbai Aug 2018- Present

  This work is an extension of my work from my master’s thesis. The work involves understanding statistical analysis of data from a cosmology perspective. It also involves using libraries such as HealPy, PyFITS to manipulate the data. During my master’s thesis, I was able to compute the cross correlation from various galaxy surveys. Currently I am trying to use the theory of Halo models to constraint various cosmological and astrophysical parameters using these cross-correlations.

• Turbulence in ICM using tSZ maps

  Along with: Rishi Khatri

  TIFR, Mumbai Jan 2020- Present

  We try to probe turbulence around the virgo cluster by using tSZ maps.

• Component seperation from Planck skymaps using unsupervised machine learning

  Along with: Rishi Khatri

  TIFR, Mumbai Feb 2019- Present

  This work was part of my master’s thesis during my time at TIFR. This work involves using unsupervised clustering techniques to perform component seperation on Planck skymaps to extract tSZ maps.

Past Research Projects

• Using Neural Networks to detect non-Gaussianities in the CMB

  Along with: Tuhin S Roy, Rishi Khatri

  TIFR, Mumbai Dec 2018- Feb 2018

  This work tried to use neural network to detect non-Gaussianities in the CMB Sky map. The work involves finding novel methods to use neural networks on a spherical manifold. We moved on to a different project involving machine learning pretty soon. I would like to get back to this soon.

• Estimation of the mass gap in modified SYK hamiltonians

  Supervisor: Gautam Mandal

  TIFR, Mumbai Aug 2018- Jan 2019

  Worked on numerically estimating the massgap in Modified SYK hamiltonians. The work involved understanding the conformal limit in the SYK model and analytically computing the massgap in the conformal limit and numerically trying to compute the eigenvalues of large dimensional matrices to get as close to the conformal limit as possible.

• Quark gluon discrimination using Deep Neural Networks

  Supervisor: Tuhin S Roy

  TIFR, Mumbai Aug 2017- Jan 2018

  Worked on building a convolutional neural network classifier to classify the quark jets from the gluon jets in particle accelerators. The work involved understanding the basics of neural networks and machine learning, build it using Tensorflow, make simulations of particle accelerators using Pythia, jet clustering using FastJet and understanding certain physics observables to classify jets the conventional way.

• Rigidity percolation in wet granular systems

  Supervisor: Purusattam Ray

  Institute of Mathematical Sciences (IMSc), Chennai Jun 2015 - Aug 2015

  Worked on understanding rigidity transition by using percolation theory and modelling it similar to jamming transition in granular systems.