A collection of machine learning, NLP, computer vision and sports analytics projects built from real datasets.
ML & Predictive Modelling
End-to-end flight status analysis and prediction using PySpark on large historical datasets, modelling delays, cancellations and other status-related outcomes.
January 17, 2024
ML & Predictive Modelling
Decision Tree algorithm to predict car acceptability based on price, maintenance, doors, boot size and safety rating, using the UCI Car Evaluation Dataset with pre-pruning and visualisation.
August 26, 2023
ML & Predictive Modelling
Analysis and prediction of heart disease using a dataset of 900+ patients, uncovering patterns to predict cardiovascular risks and support informed medical decisions.
July 01, 2023
Computer Vision & Sports
YOLO-powered entity detection in football footage with real-time tracking, 2D map representation and raw data extraction for post-match analytics.
June 12, 2023
NLP & Generative AI
Sentiment classifier built with NLTK's VADER and HuggingFace RoBERTa Transformers, applied to Amazon customer reviews to compare lexicon-based vs. deep-learning approaches.
May 15, 2023
ML & Predictive Modelling
Classification model to predict bad loans, helping banks decide whether to approve credit by accounting for risk based on historical applicant data.
April 28, 2023
NLP & Generative AI
Fine-tuned GPT-2 to generate coherent, contextually relevant Netflix show descriptions, demonstrating auto-regressive language model capabilities on a creative text task.
August 18, 2023
NLP & Generative AI
Rule-based and ML-driven chatbot built from scratch, capable of understanding user intent and generating contextually appropriate responses.
May 4, 2024
Computer Vision & Sports
Exploratory data analysis of summer transfer activity across top European leagues, using charts and graphs to answer key questions about transfer spending and trends.
June 14, 2024
ML & Predictive Modelling
Machine learning model predicting Premier League match winners using time-series data, with a focus on feature engineering, error measurement and iterative performance improvement.
September 22, 2024