suntzunami Goto Github PK

Hi
I'm a recent graduate from the Indian Institute of Technology Roorkee, where I obtained a bachelor's degree in Production and Industrial Engineering.

My current interests include data science, finance, multi-objective optimization, health research and scientific visualization. I have actively contributed to various projects in these domains, some of which can be found here.

Hobbies: Football, Hiking, Photography, Video editing, Watching anime/reading manga.

Misc. animations
Google Summer of Code 2021
I brought tools for molecular visualisation to FURY by creating a python module which can be used to create the following molecular representations: Stick Ribbon Ball and Stick Sphere Source code for the module Final report of the work done GSoC python blogs

1. Pairs Trading (Sep 2020 - Jan 2021): This was a group project where I worked with my friends in Finance Club IIT Roorkee and we implemented some pairs trading algorithms mentioned in scientific papers. I created a MSE criterion to select the pairs and visualised the returns obtained by selecting various pairs (check the figure below).
   


2. An attempt at making an algorithm that can outperform S&P 500 between 1980-2020 (Jul 2020)
   
   (Figure above compares the returns of the algorithm with S&P 500)


3. Algorithm that aims to generate profitable returns on an intraday basis (used RSI and ADX indicators to devise a profit booking logic and position management logic) (Jul 2020)
   
   (Figure above displays RSI for the given dataset)

Back in high school, I wasn't particularly fond of computer science as it seemed too abstract. This project gave me an opportunity to apply what I'd learned over the course of the subject and this was the first time that I enjoyed programming.

Investigating Innovized Progress Operators with Different Machine Learning Methods

This paper explores performance of different Machine Learning (ML) methods on Reference vector (RV) based Evolutionary Multi- and Many-objective Optimization algorithm NSGA-III utilizing innovized operators IP2 and IP3. Through an exploratory analysis of eight different ML methods, the results suggest that kNN emerges as the winner for both IP2 and IP3 in terms of the Hypervolume (HV) metric and corresponding run-time.
Paper accepted in International Conference on Evolutionary Multi-Criterion Optimization: EMO 2023 (second author) DOI: 10.1007/978-3-031-27250-9_10

The most important thing is to try and inspire people
so that they can be great at whatever they want to do.

Kobe Bryant