VARUN MACHHALE KUMAR

About me

Professional

In December 2024, I graduated from Purdue University with a Master of Science (MS) degree in Electrical and Computer Engineering. My primary interest includes applying machine learning to solve prevalent problems in neuroscience. For my master's thesis, I collaborated with Dr. Alexander Chubykin and Dr. Joseph Makin on probing the mechanism of memory encoding via theta oscillations in the mouse visual cortex. I worked on building a deep-learning model based on predictive-coding dynamics to simulate the results of a wet lab experiment. I also used mice neural data recorded using Neuropixels probe to perform stimulus and genotype classification in Wild Type and FragileX mice. Additionally, during my time in the lab, I worked on building an end-to-end neural data processing pipeline to streamline the analysis of mice neural data recorded from Neuropixels probes in the laboratory.

I received my Bachelor of Technology (BTech.) degree from National Institute of Technology Karnataka (NITK), India in Electronics and Communication Engineering in May 2021. During my bachelors, I worked on applying deep learning to segment medical images including bladder cancer and cardiac chamber segmentation. For my bachelor thesis, I worked with Dr. Hardik Pandya at BEES Lab, Indian Institute of Science (IISc) for detecting and classifying epileptic subtypes from recorded EEG signals. This led to a journal publication in the Biomedical Signal Processing and Control journal in 2023 and the granting of an Indian patent in January 2025.

Personal

I am a naturally curious learner and always eager to explore how things (including people) work, whether it’s artificial intelligence, neuroscience, the human cortex, the mechanics of everyday life, religion, politics, or the millions of brilliant minds that exist (or once did).
My hobbies are scattered. I love learning new languages —- I already know four and am slowly but surely learning a fifth. Salut, ça va bien ? I can play two minutes of two songs on the piano, made a mark with my guitar (scar marks i.e.), love cycling, and enjoy traveling to new places. To be honest, I’m not very good at sports, but I am a good sport ;) — honestly, a very competitive one that too. I play badminton and basketball and know the rules of chess just enough to defeat a "few" players. I care deeply about this planet (and the people on it) and am always looking for ways to reduce and recycle waste. We’ve got to take care of the only planet we have -- for now -- right?. Apparently, I also collect a lot of hats!? I value privacy and integrity, and strongly believe in the people and companies working to build a better internet and a better world. I’ve learned a lot through my experiences and the people around me and a day in my life is never the same as the one before, which keeps it interesting. Above all, I firmly believe in uplifting everyone.

Interests

  • Machine Learning
  • Signal Processing
  • Neuroscience
  • Neurodegenerative diseases
  • Medical Imaging
  • Computer Vision
  • Natural Language Processing

Education

  • Purdue University

August 2022 - December 2024

  • Master's in Electrical and Computer Engineering
  • National Institute of Technology Karnataka (NITK)

July 2017 - May 2021

  • Bachelors in Electronics and Communication Engineering
  • • Department Rank: 2 (out of 106 students)
  • Lourdes Central School

June 2005 - May 2017

  • AISSCE CBSE Class XII (2017)
  • • Dakshina Kannada District Topper

  • AISCE CBSE Class X (2015)

Experience

Graduate Research Assistant

Purdue University, West Lafayette
August 2023 - December 2024

Probing memory encoding via theta oscillations in deep neural networks

  • Applied principles from predictive coding theory in a deep neural network to replicate a wet lab experimental finding
  • Implemented a deep learning model that elicits oscillations in the theta frequency band (4-8 Hz) and analyzed the changes in the power to familiar vs novel images
Thesis

Graduate Research Assistant

Purdue University, West Lafayette
August 2023 - December 2024

Studying the learning impairments in FragileX mice

  • Worked on developing machine-learning based classifier models that can predict stimuli and mice genotypes using neural signals recorded from neuropixels probes
Thesis

Graduate Research Assistant

Purdue University, West Lafayette
August 2023 - December 2024

LemonPy: A neural data analysis package

  • Developed an end-to-end data processing pipeline in Python to streamline the analysis of neural data recorded by neuropixels probe in the lab

Graduate Researcher

Purdue University, West Lafayette
November 2022 - July 2023

Speech Decoding using ECoG signals

  • Devised a preprocessing pipeline that computes time-aligned phonemes from recorded speech data that can be used to train a speech decoder

Undergraduate Research Intern

BEES Lab, Indian Institute of Science
January 2021 - September 2021

Automated extraction of ECG signals from PDFs

  • Developed an algorithm to automatically convert non-overlapping 12-lead ECG signals recorded in the form of PDFs to numerical data

Undergraduate Research Intern

BEES Lab, Indian Institute of Science
July 2020 - September 2021

Automated Epilepsy Classification using EEG signals

  • Developed a fully automated epilepsy classification system to classify epileptic patients into various subtypes of epilepsy using EEG signals
Paper

Undergraduate researcher

NIT Karnataka, India
October 2020 - March 2020

Bladder cancer segmentation using deep learning

  • Implemented a deep learning neural network to segment immunopositive and immunonegative tumour cells in the bladder and calculated its Ki-67 index
Code

Summer Research Intern

CeNSE, Indian Institute of Science
May 2019 - July 2019

Small signal parameter extraction of GaN AND GaAs transistors

  • Implemented a method to extract the parasitic capacitances, inductances and resistances from the S parameters of a Gallium Nitride (GaN) and Gallium Arsenide (GaAs) transistor.
  • From the extrinsic parameters, the intrinsic parameters of the MOSFET were further extracted.
Report

Projects

Vision transformers with learnable resizer networks

Purdue University
  • Implemented and compared performance of Vision Transformer models after integrating learnable resizer modules.
  • Assessed the model on two datasets: Beans dataset (0.97 score) and a subset of CIFAR10 dataset (0.95 score).
Code Report Video

Speech enhancement using CNN-GAN

National Institute of Technology Karnataka
  • Collaborated in a team to evaluate a time-frequency mask based approach for speech enhancement through convolutional generative adversarial networks (CNN-GAN)
  • Evaluated speech quality metrics between mask (PESQ score: 2.34) and non-mask based models (PESQ score: 2.05).
Code Paper

Heart-chamber segmentation using deep learning

National Institute of Technology Karnataka
  • Designed and implemented a two-chamber echocardiogram segmentation algorithm using deep learning and compared its performance using various performance metrics with existing deep learning architectures.
  • Demonstrated improved generalization using various performance metrics over existing deep learning architectures.
Code Paper

Thirsty crow line-following robot

E-Yantra (2018), Indian Institute of Technology, Bombay
  • Through participating in the national-level E-Yantra robotics competition, we built a line following robot to pick scattered magnetic pebbles and deposit them to a deposition zone (pitcher).
  • The robot was designed to complete the task of collecting and depositing the pebbles in the shortest time possible using A* algorithm.
  • Augmented reality was used to showcase pebbles, water pitcher and the crow (robot), and updates to the water level and pebble pile was made every time a pebble was picked and dropped in the water pitcher.
Code Videos

Awards

MPOWER Scholarship 2022

MPOWER Financing

Honorary Mention Winner for the June 2022 monthly scholarship.

The MPOWER scholarship is awarded to international students studying in the US based on the applicant’s qualifications, quality of essays submitted and applicant’s potential as a scientist to make positive impact.

DAAD WISE Scholar 2020

Deutscher Akademischer Austauschdienst (German Academic Exchange Service) Working Internships in Science and Engineering (WISE)

Received the DAAD WISE Scholarship for the year 2020 to intern at a German university.

DAAD is a German organization that promotes international academic and research exchange programs. It supports students, researchers, and academics from around the world in pursuing educational and research opportunities in Germany.

The program targets around 100 students all over India pursuing a degree in the fields of science and engineering who wish to do a 2-3 months research internship at a publicly-funded German higher education institution.

Award letter

CeNSE scholarship 2019

Centre for Nano Science and Engineering Scholarship, Indian Institute of Science (IISc)

Among the 26 students throughout India to be selected for the CeNSE scholarship program in the year 2019.

The CeNSE scholarship program provides an opportunity for undergraduate and masters students from India and abroad, selected through a rigorous and competitive process, to intern at the Centre for Nano Science and Engineering, IISc for a period of 8-10 weeks.

Award letter

Institute Scholarship

National Institute of Technology Karnataka (NITK)

Every year top five meritorious students in each department are awarded the Institute scholarship from NITK.

Received the scholarship three times: Jan 2018, Jan 2019 and Sept 2021

Publications

Biomedical Signal Processing and Control Journal (2023)




Spatiotemporal analysis of interictal EEG for automated seizure detection and classification

Joshi, Rathin & M., Varun & Agrawal, Dr & Rao, Avinash & Mohan, Latika & Jayachandra, Mahesh & Pandya, Hardik. (2023). Spatiotemporal analysis of interictal EEG for automated seizure detection and classification. Biomedical Signal Processing and Control. 79. 104086. doi: 10.1016/j.bspc.2022.104086

Patents




Epileptic Seizure Classification

Rathin K. Joshi, Varun Kumar M., Megha Agrawal, Latika Mohan, M. Jayachandra and Hardik J. Pandya. 2022. Epileptic Seizure Classification. Indian Patent No.: 558139 (Filed: August 10, 2022; Granted: January 17, 2025)

Academic Courses

Machine Learning and related

Purdue University

  • ECE 69500: Generative Models
  • BIOL 56200: Neural Systems
  • ECE 57000: Artificial Intelligence
  • ECE 62900: Introduction to Neural Networks

National Institute of Technology Karnataka (NITK)

  • EE871: Machine Learning
  • EC899: Applications of Machine Learning Techniques For Medical Image Analysis
  • EC348: Image and Video processing
  • EC347: Speech and Audio processing
  • EC272: Digital Signal Processing

Mathematical courses

Purdue University

  • STAT 51100: Statistical Methods
  • ECE 60000: Random Variables

National Institute of Technology Karnataka (NITK)

  • EC457: Statistical Analysis
  • MA110: Engineering Mathematics I
  • MA111: Engineering Mathematics II (Calculus)
  • EC454: Numerical Analysis and Applications
  • EC224: Mathematics for Electronics and Communication Engineering

Entrepreneurship

Purdue University

  • ECE 69500I2I: Ideas To Innovation Part 1
  • ECE 69500CEL: Communication for Engineering Leaders
  • ENTR 50000: Technology Realization Seminar

Get in touch