Greetings! I'm Mahule Roy, currently immersed in the captivating worlds of Metallurgical and Materials Engineering at the National Institute of Technology Karnataka, Surathkal. Envision me as a dynamic fusion enthusiast, relentlessly pursuing the frontiers of knowledge across material science, quantum physics, biomedical wonders, and the enchanting realm of Artificial Intelligence.
My intrigue with the vast landscape of knowledge first ignited during my second year when I embarked on Andrew Ng's transformative Coursera Machine Learning course. Since that enlightening encounter, my passion for exploration has known no bounds. I find myself endlessly fascinated by the intricate dance between algorithms and reality, continually seeking avenues to expand my horizons.
Beyond the confines of traditional academia, my journey unfolds through a tapestry of diverse projects that reflect my insatiable curiosity. From crafting intelligent systems and delving into the intricacies of robotics to exploring the creative realms of generative AI and conjuring solutions in mesmerizing image processing spectacles—each project becomes a canvas for innovation and a testament to my boundless interests.
Venturing beyond the theoretical, I've accumulated valuable work experiences in IT, R&D, Machine Learning, quantum realms, and the manufacturing cosmos. This diverse exposure has shaped me into a polymath of sorts, ready to tackle challenges with a multifaceted approach.
Currently, I'm orchestrating a harmonious blend of Material Science and ML, immersing myself in the symphony of data and molecules. This unique fusion has propelled me into groundbreaking pursuits, from enhancing drug delivery mechanisms to uncovering superior materials that defy conventional limits.
In the cosmic dance of knowledge, my interests span across almost everything, embracing the intersections of technology and science with unbridled enthusiasm. This insatiable curiosity, coupled with my extensive project portfolio and diverse work experiences, positions me as a versatile innovator capable of navigating the complexities of intelligent systems and driving interdisciplinary projects in academia and industry.
Paper on Machine Unlearning methods
Poster presentation on TransMedTech 2024: International Conference on Advances and Challenges in Medical Technology Translation
Selected in INTERNATIONAL JOURNAL OF CURRENT SCIENCE. This paper explores the application of OpenMP for accelerating the training of neural networks in Fashion MNIST data recognition.
Selected in more than 4 international conferences for presentation. This paper aims to provide a clear understanding of the current state of QML research and its potential impact on future computational capabilities.
A comprehensive analysis is conducted using traditional ML models, including Support Vector Machine (SVM), Random Forest, and Logistic Regression, alongside advanced DL models employing Convolutional Neural Networks (CNN).
Rockwell Hardness C Test was performed on a sample of steel provided and the experimental readings are recorded with inclusion of theory, procedure, pictures of the samples after testing is done.
Description: Selected to attend the MENA ML International Winter School organized by Google DeepMind, QCRI, and HBKU, while presenting my paper on Machine Unlearning as a poster
Description: Chosen from 500+ applicants to attend this workshop, featuring keynote speakers from top U.S. universities, including Prof. Dimitri Bertsekas (ASU, MIT)
Title: Quantum Machine Learning: The Superhero That Classical Machine Learning Never Knew It Needed
Description: Presented paper on Quantum Machine Learning: The Superhero That Classical Machine Learning Never Knew It Needed.
Outcome: Received positive feedback for innovative methodology and practical implications for industries like quantum.
Title: Effects of Screen Time on Glaucoma
Description: Highlighted the role of AI in healthcare, particularly in improving early-stage glaucoma detection. Collaborated on research involving screen time analysis and vision health monitoring.
Outcome: Paper featured in the top 10 innovative projects of the year in medical technology.
Description: Attending this international conference at CDS Dept, IISc.
Title: Comparative Study of Machine Unlearning Techniques for Computer Vision and NLP Models
Description: Presented paper on Machine Unlearning is an emerging field in machine learning that focuses on efficiently removing the influence ofspecific data from a trained model.
Outcome: Will publish in AI Expo Proceedings, receiving citations from researchers in sports analytics.
This repository contains my implementations of agents with abstractions for various tasks.
This project implements an Image Caption Generator using Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) in Python.
Welcome to the AI Toolbox Agent project! This project aims to develop an artificial intelligence agent capable of utilizing a wide range of tools provided to it, enabling versatility and adaptability across various tasks and domains.
Welcome to the conversational chatbot project utilizing HuggingFace's Zephyr 7B Alpha model. This README will guide you through setting up and utilizing the chatbot effectively.
A chatbot based on sklearn where you can give a symptom and it will ask you questions and will tell you the details and give some advice.
LORA is a technique for adapting large language models using low-rank approximation, implemented using PyTorch. This project contains the code for applying LORA to the MNIST dataset, as well as explanations and examples to help you understand and use the technique effectively.
Welcome to the project dedicated to exploring the fascinating world of Joint Embedding Predictive Architectures (JEPAs) and self-supervised learning (SSL) based on the paper Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecture
This is the final project of AI Programming with Python Nanodegree program by Udacity
In this project, I utilized three RL algorithms to teach our agent to walk which were Q-learning, Deep Q-Network (DQN), and Twin Delayed DDPG (TD3).
Welcome to the Mars Lander Rover project! This simulation allows users to control a rover landing on Mars, implementing basic physics and controls to safely land on the Martian surface.
Welcome to the Mario Game! This is a simple 2D platformer game developed using Python. The game features classic Mario gameplay elements such as jumping, collecting coins, and avoiding enemies.
This project implements a Snake AI game where an AI-controlled snake competes against a user-controlled snake. The goal is to grow the snake by consuming food while avoiding collisions with walls and itself.
This project illustrates the steps taken towards optimizing this simple chess engine in an incremental way, starting by parallelizing it's search algorithm, then applying some optimization techniques, and finally trying out the iterative deepening variation of the classical minimax algorithm.
Demonstrated exceptional performance, achieving an average waiting time reduction of 30%. Journey time improvement of 25% compared to baseline simulations
95% for player detection & 90% for ball detection. Detection accuracies of 92% for players, 85% for referees, and 88% for footballs
Developed a responsive chatbot capable of handling prompts like "Hello", "How are you?", and "Bye". Displaying motivational quotes & implemented various features
This application is entitled School Registration System. This is a simple web-based application developed in Python and Django Framework.
This project is an interactive shell providing the user to borrorw, return,donate books from the library system.
Got 20% reduction in waiting times. 15% decrease in journey times compared to baseline simulations
A casual 3D game consisting on moving a ball towards the finish line while trying to avoid the obstacles on the way.
Welcome to the Capstone Project on Machine Learning - Recommendation System, as part of the IBM Coursera course.
This project contains the 4 week AI Capstone project assignment using keras as a part of IBM-AI Engineering
roymahule26@gmail.com
