Projects

I created an AI that would play a game of chess. The AI is trained to learn to calculate the best position 5 depths in based on the parameters I have asked it to check. The image is one of the games I played, as white, against my own AI. Simultaneously, my app will print out the notation of the pieces up to the first 58 moves on both sides so that players can track their moves. This engine has proved to be around an estimated ELO of 823. I tried and tested the ELO against chess.com bots to get an accurate approximation of the ELO. One of the key algorithms used in the development of this engine is the alpha-beta pruning algorithm, which reduced my "thinking time" by 150%. To further improve this engine, the best plan of action would be to track each game, and reanalyze such that it can learn at which move did the evaluation go against it, so that it can learn from its mistakes and slowly fine tune itself.

Using python jupyter Notebooks and kaggle, I trained a model that would predict the house value of a property in California, given some parameters. The model takes into account the proximity to water, amount of bedrooms and other factors when calculating the value. This project is an introduction to model training as I have an interest in Machine Learning and using predictive AI to impact the future. The model currently runs at an 87% accuracy and uses Python modules like scikitlearn to evaluate and analyze the features.

Using a football API, which returns the complete league table of an indicated year, and a weather API, my app returs the standings of a selected year. Then, using that response, it will return the city which hosts the first place team of that league and returns a full analysis of the current weather of that city. The app uses a dropdown menu of the top 5 soccer leagues in Europe: English Premier League, Ligue 1, Bundesliga, La Liga, and Serie A and another menu of the years from 2012-2022. Once those selections have been made, then the football API will retrieve the league standings of the specified league at the specified year. After doing so, it will use the weather API to return the current weather of the area of where the city is located.