"I am a Ph.D. student and a Graduate Research Assistant focusing onReinforcement Learning and Multi-Armed Bandits.My ongoing Ph.D. research explores advanced algorithms and neural forecasting models to improvethe quality of the decisionstaken by Multi-Armed Bandits and make them robust against data integrity issues.Practical applications include automotive vehicles, addressing networking challenges, improving adaptive cyber-physical systems and even stock trading. I am committed to advancing AI and machine learning by integrating practical engineering skills withcutting-edge academic research to
address complex challenges and drive innovation."

I prioritize client collaboration, encouraging open communication

I'm adaptable to time zone communications

I constantly try to improve

My tech stack

Tech enthusiast with a passion for development.

The Inside Scoop

Currently building a Multi-Armed bandit framework

Do you want to start a project together?

Educational Qualifications

2015 - 2019
B.E. in Computer Science and Engineering
Graduated from Jadavpur University, India with
8.37(/10) CGPA and Honours First Class
2021 - 2022
M.S. in Computer Science
Graduated from Northeastern University, Boston with
3.67 GPA
2023
Currently pursuing Ph.D. in Computing and Information Sciences
@Rochester Insitute of Technology, Rochester

My work experience

Latest News

One Workshop Paper and One Poster accepted at GECCO 2024 Conference

A few of my research publications

[1]	Roy, Soumya Deep, et al. "Computer aided breast cancer detection using ensembling of texture and statistical image features." Sensors 21.11 (2021): 3628.
[2]	Sanyal, Ritabrata, Devroop Kar, and Ram Sarkar. "Carcinoma type classification from high-resolution breast microscopy images using a hybrid ensemble of deep convolutional features and gradient boosting trees classifiers." IEEE/ACM transactions on computational biology and bioinformatics 19.4 (2021): 2124-2136.
[3]	Kar, Devroop, et al. "Fuzzy mutation embedded hybrids of gravitational search and Particle Swarm Optimization methods for engineering design problems." Engineering Applications of Artificial Intelligence 95 (2020): 103847.
[4]	Kar, Devroop, Asif Iqbal Middya, and Sarbani Roy. "An approach to detect travel patterns using smartphone sensing." 2019 IEEE international conference on advanced networks and telecommunications systems (ANTS). IEEE, 2019.

A small selection of recent projects

iCMAB: Augmenting Bandit Strategy with EXAMM for Stock Trading

Using EXAMM to build RNN models to predict future returns on stocks and Bandits to drive trade decisions.

Check Repository

/ui.goto.link

EzFlow: A Pytorch library for Optical Flow

Collaborated on finishing an open PyTorch library for Optical flow tasks, allowing users to design and train 6 highly configurable models like RAFT, PWCNet, VCN, etc. on 9 supported datasets like FlyingChairs, KITTI, Sintel, etc.