I am a 5th year PhD Candidate in the Department of Chemical and Biomolecular Engineering at University of Notre Dame. My research is in the area of computational chemistry where I parametrize Density Functional Theory (DFT) calculations to build mathematical models. These models are used to perform Monte Carlo simulations and molecular dynamics.
Driven by my research needs, I started learning about modelling algorithms and techniques and it was while exploring concepts of feature selection for my research, I came to know about the field of Data Science. I found the field fascinating and read books and pursued specialization courses on Coursera, edX, etc., to hone my knowledge and skills on the topic. I am looking forward to pursuing a career that involves exploration of datasets, making visualizations and solving data intensive real world problems.
During my stint with IBM Research Lab, I worked in collaboration with Australian Fashion Designer, Jason Grech, to develop a web application that makes recommendations about the upcoming trends in color and print patterns in the fashion industry.
In the project, I wrote python scripts that crawled fashion and social media website to download more than 500k images. The images came from various sources like celebrity instagram profiles, collections by designers in various fashion weeks over the last 15 years, etc. With the images, we extracted the colours and print patterns and build a prediction model using Gaussian process regression to predict the colors that are likely to be popular in the future seasons.
The complete analysis was presented as a web application that the designer used to see the color patterns and trends that were popular over the last 15 years. Based on the recommendations provided by our application, Jason presented a couture collection at Melbourne Fashion Week' 16.
In this position, I studied various reaction enhancing techniques like sonication, electro-organic synthesis, and microphase assistance and carried out experiments to optimize Crystal Size Distribution (CSD) of reactant particles for enhancing reaction rate in heterogeneous systems.
I developed a software program to predict the optimum particle size of the reactants and the scope of microphase assistance in a specified reaction system.
This is a website that I designed to learn about the web development and use of d3.js. The javascript library d3.js is a tool to make visually appealing data driven documents.
Currently, the website provides two kinds of analysis.
In this Data Science Specialization Capstone project, I developed a text prediction program which uses n-grams up to 4-grams to predict the next word for a given phrase using a simple Back Off algorithm. The model is trained on a dataset that comprises twitter feeds, blogs and News articles in English. The script for the program is written in R using R studio and the web rendering is provided by shinyapps. The scripts and other details of the project can be accessed here.
Add all the kaggle project here.............
Following is a list of the online courses and certifications that I have completed: