Finance and Data Science Projects Portfolio

Welcome to my portfolio of finance and data science projects! This repository contains a collection of my most significant work in the fields of machine learning, data science, and financial modeling. Each project is documented with the relevant code, datasets, and explanations to help you understand the scope and impact of the work.

Projects Overview

1. Predicting Housing Prices

Overview: Built a linear regression model to predict housing prices in Cook County using scikit-learn, achieving an RMSE of less than 240k on test data.

• Key Technologies: Python, scikit-learn, Pandas
• Highlights: Enhanced model accuracy with advanced feature engineering. Addressed regressive tax implications.
• Link to Project Files

2. Multi-Agent Search Algorithms

Overview: Designed and implemented multi-agent search algorithms, including Minimax, Alpha-Beta Pruning, and Expectimax, for the classic Pacman game.

• Key Technologies: Python, AI, Search Algorithms
• Highlights: Optimized decision-making in Pacman, achieving a high win rate against multiple ghosts.
• Link to Project Files

3. Ghostbusters: Inference with Bayes Nets

Overview: Implemented exact and approximate inference algorithms using Bayes Nets for tracking and capturing invisible ghosts in Pacman.

• Key Technologies: Python, Bayesian Inference, Particle Filters
• Highlights: Real-time tracking with high accuracy using particle filters and sensor data.
• Link to Project Files

4. Reinforcement Learning for Game AI

Overview: Implemented and optimized reinforcement learning algorithms to train an AI agent for Gridworld, Crawler, and Pacman games.

• Key Technologies: Python, Reinforcement Learning, Q-learning
• Highlights: Achieved a 90% win rate by applying advanced learning strategies.
• Link to Project Files

5. Machine Learning for Digit Classification and Language Identification

Overview: Built and optimized neural networks for tasks like digit classification and language identification using PyTorch.

• Key Technologies: Python, PyTorch, Neural Networks
• Highlights: Achieved a 97% accuracy on digit classification and 81% on language identification.
• Link to Project Files

Getting Started

To explore these projects, you can clone the repository and navigate through the individual project directories. Each project contains:

• Source Code: The main code files used to implement the project.
• Data Files: Any datasets used, if applicable.
• Documentation: Detailed explanations of the project objectives, methodologies, and results.
• Instructions: How to run the code, reproduce results, and any dependencies required.