Shreyas Kowshik
shreyaskowshik@gmail.com

I am a final year Integrated Msc. student majoring in Mathematics and Computing at Indian Institute of Technology Kharagpur with a Micro Specialisation in Optimization Theory and Applications.

I have been fortunate enough to work under the guidance of some amazing mentors. Currently I am working with Prof. Buddhananda Banerjee on Statistical Changepoint Detection in Time Series for my Master's Thesis. In the past I have worked with Prof. Guy Van den Broeck and Yitao Liang on Tractable Probabilistic Modelling at the UCLA StarAI Lab. I have also spent some awesome time as an Applied Scientist Intern at Microsoft working on Automatic Speech Recognition and Google Summer of Code on open-sourcing machine-learning paper implementations in Julia.

Outside of work (contrary to the general belief, I do have a life :D) I enjoy playing the keys, guitar, and racket sports especially ping-pong and badminton. I am a hard core rock music fan and horror movie buff. I do get fascinated by things written well, be it beatiful poetry or well structured code.

CV  /  Github  /  LinkedIn  /  Twitter

profile photo
Updates

  • [June 2021]  Paper on Stronger Ensemble Learning of Probabilistic Circuits accepted at TPM Worshop, UAI'21
  • [May 2021]  Started an Applied Scientist Internship at Microsoft IDC
  • [April 2021]  Second runners up in shared task in NLP4IF workshop, NAACL'21. Paper accepted!
  • [May 2020]  Began remote research internship at StarAI Lab at UCLA
  • [March 2020]  Began remote internship at Computer Vision Lab at Linkoping Uniersity
  • [May 2019]  Selected for Google Summer of Code under The Julia Language
  • [December 2018]  Paper on Deep Learning for Traffic Sign Classification accepted at ICPRAM
  • [July 2017]  Started my journey at IIT Kharagpur as an undergraduate

Research

I am interested in Machine Learning in general, but currently focussing on Probabilistic Machine Learning and Bayesian Inference. I aim to bring insights from a statistical and algorithmic viewpoint to build systems that work in the real world. This amounts to building systems that can scale well to large dimensions and data while understanding when they are going wrong. Tools from Probability Theory and Statistics can help answer such questions. I am also interested developing robust deep learning algorithms for real world-applications like autonomous-driving, robotics and healthcare.

Testing for a Change in Mean of a Weakly Stationary Time Series
Guide: Prof. Buddhananda Banerjee, IIT Kharagpur
Thesis  /  Slides

Changepoint testing is the detection of an abrupt change in the distribution of a time series. Change in mean detection is generally posed as a hypothesis testing problem. Self-Normalizing statistics aid such a detection by being adaptive to the presence and location of a changepoint. We propose such a statistic that acheives sharper power rise compared to previous approahces on deviation from the null-hypothesis.

Independence Based Learning of Probabilistic Circuit Ensembles
Guide: Prof. Guy Van den Broeck, Yitao Liang, StarAI Lab, UCLA
Paper  /  Slides  /  Poster  /  Code  /  TPM Workshop, UAI'21

Structured-Decomposable Probabilistic Circuits model the data by encoding different types of context-specific-independences (CSIs). One can use these CSIs as prior information for initializing the mixture components of an EM-ensemble of circuits. We propose an algorithm by partitioning the data in coherence with these CSIs and training a circuit on each of them to get a stronger initialization of an ensemble.

Graph Representation Learning for Road Networks
Guide: Prof. Michael Felsberg, Zahra Gharaee, Linkoping University
Paper  /  Pattern Recognition Journal

Real world road networks can be represented as a graph with intersections as nodes and roads as edges. One can use graph-based deep learning approaches to learn efficient representations of such nodes. We survey various algorithms in the literature for this task and propose a DFS based aggregation scheme to aggregate information from nodes far away from a given node.
Predicting Binary Properties of Tweets : NAACL Shared Task
Paper  /  Code  /  Task Link  /  NLP4IF Workshop, NAACL 2021

Tweets during the time of a pandemic, if contain false information, can lead to spread of misinformation. The task was thus to automatically answer multiple binary questions (in yes/no) pertaining to a tweet. This problem was framed as a multi-output learning problem with answering of each question formulated as a task. Inter-task attention modules were proposed for aggregating information on top of a large-language-model. Approach resulted in runners-up position.

Perception for Autonomous Driving
Guide: Prof. Debashish Chakravarty, Autonomous Ground Vehicle Research Group, IIT Kharagpur
Report  /  Video  /  Code

Perception systems are crucial for robots to make sense of their surroundings. Developed an end-to-end object-detection and tracking pipeline for traffic-sign detection and classification for Indian Signs. Built a real-time road segmentation newtork for road-region segmentation. Both modules tested on electric vehicle Mahindra-e2o.

Key Academic Projects
Uncertainty Estimation in Deep Learning
Course: Advanced Machine Learning, Prof. Pabitra Mitra
Report

Surveyed multiple papers on calibration and uncertainty in deep neural networks. Methods like MC-Dropout, Bayes by Backprop, SGLD and p-SGLD were implemented and tested on MNIST. Uncertainty estimates of the predictions were obtained on out-of-sample images as a proof of concept.

Machine Learning for Portfolio Optimization
Course: Optimization Methods in Finance, Prof. Geetanjali Panda
Report

Built forecasting models to predict future stock prices of 30 stocks using LSTMs, Linear Regression and Support Vector Regression. Forecasts were used to select top `N` performing stocks to build a portfolio. Markowitz models were applied for portfolio optimization. Use of forecasts lead to better sharpe-ratios and returns in general on the test-set.

Regression Analysis for Medical Cost Estimation
Course: Regression and Time Series, Prof. Buddhananda Banerjee
Report

Built statistical linear regression models to predict medical costs of patients. Improved R-squared value of the fit by implementing and analysing residuals-vs-fitted, sclae-location, residuals-vs-leverage and normal-qq plots.

Google Summer of Code
Blog

Implemented multi-modal image2image translation GANs like pix2pix, cycleGAN SRGAN purely in The Julia Language. Wrote a library for reinforcement-learning containing PPO and TRPO implementations. Implemented Neural-Image-Captioning in a pure julia codebase.

128-bit AES-Feistel Encryption
Course : Cryptography and Network Security, Guide : Prof. Sourav Mukopadhyay
Code

Implemented a hybrid block cipher for encrypting inputs of length 128-bits. The first 5 rounds are AES followed by 5 rounds of a custom Feistel Cipher with variable number of blocks. Finally 5 rounds of AES follow.

Autonomous Navigation Robot for IGVC'18&19
Intelligent Ground Vehicle Competition Entry from IIT Kharagpur, Prof. Debashish Chakravarty
Report  /  Video

Built the planning and perception stack for the Autonomous Navigation Challenge at Intelligent Ground Vehicle Competition'19, Michigan University, USA. The robot had to follow a set of GPS waypoints autonomously, remaining within lane boundaries and avoiding obstacles. Used SVMs for object detection and super-pixel clustering for lane segmentation. Entry bagged runners-up position and won the early-qualification round.

Of course I did not design this page,
template adopted from here