Omkar Chaudhari

Omkar Chaudhari
Click on the email to unscramble.
Academic Email:
Personal Email:

~ Resume | Google Scholar | Github | LinkedIn ~

I am currently a PhD student in the department of Chemistry at UIUC under Professor Hirata So. I completed my Bachelors in Chemical Science and Technology from IIT Guwahati in 2024.

Please feel free to check out my resume and drop me an email if you want to chat with me!

Dec '24	Joined the Hirata Lab at the University of Illinois Urbana-Champaign as a PhD student.
Aug '24	Started as a PhD student in the Department of Chemistry at the University of Illinois Urbana-Champaign.
May '24	Completed my Bachelors in Chemical Science and Technology from IIT Guwahati.

Indian Institute of Technology Guwahati
Bachelor of Technology in Chemical Science and Technology
Nov' 20 - 24 (expected)

Student Societies:

General Secretary, Chemical Science and Technology Association (CheSTA)
Ex - Overall Projects Manager, Coding Club IIT Guwahati
Ex - Core Team, Game Development and Esports Club IIT Guwahati
Branch Representative, Chemistry Department, IIT Guwahati
Member, Aquatics Club, IIT Guwahati
Ex - Workshops Manager, Techniche, IIT Guwahati
Ex - Events Executive, Alcheringa, IIT Guwahati

A review of ensemble learning and data augmentation models for class imbalanced problems: combination, implementation and evaluation

Azal Ahmad Khan, Omkar Chaudhari, Rohitash Chandra* [paper]

In this paper, we present a computational review to evaluate data augmentation and ensemble learning methods used to address prominent benchmark CI problems. We propose a general framework that evaluates 10 data augmentation and 10 ensemble learning methods for CI problems. Our objective was to identify the most effective combination for improving classification performance on imbalanced datasets. The results indicate that combinations of data augmentation methods with ensemble learning can significantly improve classification performance on imbalanced datasets. These findings have important implications for the development of more effective approaches for handling imbalanced datasets in machine learning applications.

*Corresponding Author

Research Internship | Central Food Technological Research Institute, Mysore
Council of Scientific & Industrial Research
Dec '22 - Present

Advisor: Dr Usharani Dandamundi.

Studying graph representations of FAD Synthetase, performed network analysis and shortest path analysis of the apo and bound forms to identify topologically important residues to link the N-ter and C-ter domains.
Trajectory analysis using MDAnalysis, Dynetan, measuring key properties of protein network across snapshots such as edge between- ness, node centrality, and length of shortest path between active site residues of N-ter and C-ter Domains.
Studying Inhibitor Binding of Isomers of Expoxides of Natural Fatty acids to EPHX-2 enzyme and studying different binding poses in Autodock Vina.

Research Internship | UNSW Sydney
Aug '22 - Present

Advisor: Dr Rohitash Chandra.

Evaluation of 100 combinations of Data Augmentation and Ensemble techniques on datasets with varying imbalance ratio.
Ranking performance of combinations for binary and multi-class datasets according to F1 Score, and AUC for Binary and F1 Score averaged with Macro and Weighted Strategy for Multi Class Datasets.

Research Internship | National Chemical Laboratory, Pune
Council of Scientific & Industrial Research
Jun '22 - Jul '22

Advisor: Dr. Amol Kulkarni.

Did quantification of reaction progress in a screw reactor by calculating color percentages in images taken using Machine Learning.
Conducted literature review of Scaling up of Confined Impingement Jet Reactors for applications of nanoparticle generation and effective mixing of reactants in manufacturing process.
Studied the effect of various parameters such as Reynolds Number, Nozzle Diameter, and chamber size on the mixing efficiency of the reactor.

	Herbal Waste Management Supervisor: Prof. Ajay Kalamdhad [Proposal] Emami Pacharia Plant, North Guwahati Researched sustainable solutions for managing 1 ton of herbal waste produced per day by Emami FMCG plant. Characterisation of waste by quantitative analysis to check paramaters like BOD, TKN, TGA, ammonia and nitrate estimation. Proposed research insights to use waste to produce Briquettes, Vermicomposting by rotary drum composters, and pyrolysis to produce BioChar of calorific value around 4300 kcal/kg.
	Modification of the Job Scheduling Problem [code] [Metaheuristic Approach Report] [Heuristic Approach Report] Computer Aided Optimization Course Project Modified the Job Scheduling Problem to remove start time from the decision variables and reducing the search space. Assigned machine number to different orders using metaheuristic techniques such as Genetic Algorithm, Particle Swarm Optimiza- tion, Ant Lion Colony Optimization and a deterministic method optimized start times for orders. Multi-Objective Optimization done using the eps-constraint method to optimize cost, makespan, and production time.
	Oil Mixture Identification Supervisor: Prof. Usharani Dandamundi [code] Developed scripts to Identify Oil Compositions upto 5 different oils according to Fatty Acid standards from FSSAI. The python scripts take inputs of Fatty Acid Values from a mass spectrometer, compares them to the permitted values of 20+ oils, and outputs an Excel Sheet containing possible compositions of the Oil Mixture.
	NIR Spectra Classification Supervisor: Prof. Usharani Dandamundi [code] Classified different Millet Varieties from various Locations using Principal Component Analysis. Combined Principal Component Analysis and Linear Discriminant Analysis to boost accuracy of separating the varieties.
	Groww Simplee Route Optimization Problem [code] (Report in Github as well) Inter IIT Tech Meet 11.0 Employed Metahueristic Techniques to arrive to solutions for the NP-Hard problem Capacitated Vehicle Routing Problem with Mixed Pickup and Delivery (CVRPMPD). Created and Designed special constraints and penalties to lead the algorithm to certain solutions, considering parameters such as Location Promiximity, Delivery Dates, Capacity of Delivery Vehicles etc. Implemented the sweep clustering algorithm to break down the search space and solve sections individually, reducing calculation time from 300 seconds to 60 seconds, adding scalability to the solver to incorporate orders for large cities. 3D Bin Packing was implemented for optimal packing of the deliveries in the rider’s bag.
	Campus Rush Download Link Trailer Game Development Club, IIT Guwahati Created a 3D model of IITG in Blender, along with Multiplayer Functionality and Online Leaderboard using PHP APIs. Used UV Unwrapping to accurately texture models of various buildings of the college, modifying texture parameters like roughness, diffusion, and normal maps to give textures a realistic look while keeping performance requirements low.
	Reinforcement Learning Bot for Rocket League [code] Personal Project Created a Rocket League game bot by leveraging Reinforcement Learning Algorithms using the RLBot library. Used reward functions for Ball Velocity, Car Proximity to other Players and the Ball, and modified 8 agent parameters such as throttle/brake, steer, roll, jump/boost accordingly.