I leveraged state-of-the-art reinforcement learning techniques, including Deep Q-Networks (DQN) and Deep Deterministic Policy Gradients (DDPG), to empower a robotic system for autonomous top-down object manipulation.

Curating a comprehensive dataset for pretraining by scraping Spanish content from 120 websites, focusing on political conflict and violence across various news categories.

Improved the traditional 9 Men’s Morris game through AI modifications, incorporating advanced strategies like MiniMax and AlphaBeta pruning. These enhancements optimize decision-making, adding strategic depth and intelligence to the game.

Applied machine learning techniques to a substantial 400,000-record dataset from the 2020 CDC survey. Employed logistic regression, Naive Bayes, Decision Tree, XGBoost, and K Nearest Neighbors models to predict heart disease, aiding medical care by identifying key risk factors.

Created SADVR, a multiplayer VR game using Unity, inspired by the movie 'Gravity', to revolutionize astronaut training, providing a cost-effective and immersive platform for space exploration scenarios. This project aims to enhance safety and teamwork, leveraging the latest VR technology to simulate realistic space environments.

Conducted Sentiment Analysis on Twitter data employing a varied range of machine learning classifiers, including SVM, Maximum Entropy, Random Forest, and Naïve Bayes. Tracked and assessed public sentiment during pivotal events like elections, offering valuable insights into the dynamic landscape of social media reactions.

Addressed the challenge of enhancing review rankings on e-commerce platforms, focusing on content quality. Utilized NLTK for data preprocessing, created a pandas dataframe from Amazon.in reviews, and implemented a Recursive Neural Network (RNN) with LSTM layers to model dynamic temporal behavior for improved ranking accuracy.