Projects

These are some major projects I have done/part of after I joined IIT Bombay.

Computational Genomics

Did the project as a part of the Technical Summer School in the summers after the first year. I Learnt molecular biolog, DNA sequencing methods, sequencing algorithms involving de bruijn, overlap graphs, etc. , and other tools like BLAST, Biopython.

Modernising Land Records

Project under Prof. Milind Sohoni, that I did in my 3rd semester. Worked in a team of 4 to generate village legal-survey maps using satellite data provided by Google.

Algorithmic Trader

Developed a program for simulating the market with the help of parallel computation and socket programming. Developed and implemented efficient algorithms in C++ to find out arbitrages, approximate arbitrages, etc., under different constraints for trading. Multiple traders were run over the same market and different parameters were studied.

Computer Vision for Driverless Vehicles

SOC project did during the summer vacations after 2nd sem. Implemeted a lane detection model using the OpenCV Python library and custom trained the YOLOv5 object detection on a dataset of 8k+ images to detect road objects like potholes, cones, etc

Graph Theory

Reading project as a part of SOS (Summer of Science) that I did in the summer vacations after 1st year.

Simulating Cache Replacement Policies and Prefetchers

Implemented and simulated on ChampSim, multiple basic cache replacement policies like LRU, FIFO, LFU, etc. Read the paper - ’Adaptive insertion policies for high performance caching’ and implemented the DIP and BIP policy. Learnt the IP-stride and stream prefetchers and simulated and studied their performance on Champsim. Achieved a speedup of 1.2 on the stream prefetcher compared to the next line prefetcher.

Line Following Cargo Bot

Created an infrared sensor based line follower bot using Arduino microcontroller in a team of 6 members using AutoCAD, Fracktory and Arduino IDE.

A deep dive into ML

Implemented Anomaly detection to identify potentially anomalous servers. Implemented a virtual lunar lander with deep reinforcement learning techniques, demonstrating successful moon surface landings.

Compression and Autocomplete

Implemented DEFLATE, a fusion of the Lempel-Ziv’77 Algorithm and Huffman encoding and Engineered an autocomplete functionality that generates context-aware word recommendations.

Web-crawler

Developed a Python Web Crawler to recursively search and index website content up to a specified threshold.

Address

H9-155
IIT Bombay
Mumbai, 400072
India