Archives

An analysis of optimal Blackjack strategies using recursive decomposition and dealer probability analysis.

A technical implementation of the Actor-Critic algorithm to solve dynamic pricing. This project benchmarks RL against theoretical optima to test its effectiveness in combinatorial optimization.

A geospatial project using the Postcodes.io API and Folium to create an interactive map of Scottish schools, visualizing the relationship between pupil density and local deprivation levels.

A data-driven analysis of 2,431 Scottish state schools using K-Means clustering to categorize local authorities by pupil density and SIMD deprivation scores to identify areas requiring NGO support.

A comprehensive guide to building a functional web crawler in exactly 100 lines of Python code, covering URL success handling, depth thresholds, and data storage.

A comprehensive analysis of 160,000 visitor clickstreams using Python. This project evaluates marketing campaign performance by calculating conversion, drop-out, and bounce rates across different traffic sources and platforms.

An independent project using Python and k-Means clustering to classify over 10,000 FIFA 23 players into distinct performance segments based on Kaggle data.

A data-driven analysis of Companies House records to identify emerging UK tech hubs outside of London. This project uses custom API scraping and SIC code filtering to rank cities by tech firm density and growth.

An A-grade University of Edinburgh project utilizing Python and NLTK to analyze one exabyte of evidence from South Africa's Zondo Commission, identifying the most frequently mentioned publicly traded companies.