Algorithmic AI player for a modified “Dots, Boxes, and shape

• Spearheaded the creation of “Dots, Boxes, and Shapes,” a strategic variation of the classic game, implementing grid pattern with a unique scoring system based on shape captured. Intuitive UI and interaction method was also designed for this project.
• Developed and analyzed heuristic functions and legal move identification algorithms for varying difficulty levels in the game, achieving efficient time complexity. Utilized Network X features and alpha-beta pruning to optimize gameplay against computer opponents, demonstrating proficiency in Python and algorithmic thinking.

 NetworkX, Minimax Algorithm with alpha-beta pruning, Matplotlib.

• Made Various Deep Learning, Neural Network, and LLM classification models for building and testing supervised and unsupervised models that can recognize topics in reviews and assign individual sentiments for them. The best accuracy of 86% was achieved. 
• shop-wise and overall sentiment weights were obtained for various aspects of a café business, and recommendations were made for focus on location and facilities, aesthetics, environment, taste, and menu options.

Python,  Nvidia CUDA acceleration toolkit, Bert LLMs, ML, Neural Networks, TensorFlow, PyTorch, NLTK