I am a highly motivated Computer Science graduate currently expanding my expertise in Artificial Intelligence at Manchester Metropolitan University. My academic and research focus has deepened into advanced deep learning architectures, specifically Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Reinforcement Learning. I am particularly passionate about Medical Neural Architecture Search (MedNAS), exploring how AI can optimise medical diagnostics.
🤖 Artificial Intelligence & Machine Learning (Click to Collapse/Expand)
|
Directed an end-to-end supervised learning investigation for Pneumonia detection using X-ray feature data. Implemented rigorous preprocessing (median imputation, scaling) and conducted a comparative performance analysis between Decision Trees, k-Nearest Neighbors, and Naive Bayes to optimize diagnostic sensitivity.
|
|
Engineered an autonomous Reinforcement Learning Controller (RNN) to design CNN architectures for medical imaging. The 'Masked' agent integrates domain-specific validity rules, achieving 96.7% accuracy on MedMNIST while reducing parameter count by 300% compared to Random Search baselines.
|
|
Constructed a robust regression pipeline to predict vehicle market values, achieving an R² of 0.90. Implemented Target Encoding for high-cardinality features and optimized a Random Forest Regressor via Grid Search, reducing Mean Absolute Percentage Error (MAPE) to 15.58%.
|
🎮 Game Development (Click to Expand)
|
Engineered a 2D action-adventure platformer in Godot. Designed a robust Finite State Machine (FSM) for enemy AI (Idle, Chase, Attack) and implemented a precise 4-hit combo system using Area2D Hitbox/Hurtbox logic. Features include KinematicBody2D physics, parallax background scrolling, and signal-driven UI state management.
|
|
A dedicated sandbox for iterative mechanics testing and rapid prototyping. Focuses on modular implementations of character controllers, environmental physics interactions, and asset integration workflows. Served as the technical testbed for core mechanics (jumping gravity, collision layers) prior to deployment in Sword Scourge.
|
💻 Computer Science Fundamentals (Click to Expand)
|
Engineered a robotic navigation simulation in a dynamic grid world. Implemented and benchmarked BFS, DFS, Uniform Cost Search (UCS), and A* algorithms. The project features a custom GUI for real-time visualization and includes comprehensive batch testing to analyze path costs, node expansion rates, and time complexity.
|
|
Implemented a library of advanced static and dynamic data structures (Trees, Graphs, Hash Maps) focusing on algorithmic efficiency. Conducted rigorous Big-O time complexity analysis for sorting and searching operations. Utilized Generics to ensure type safety, code reusability, and maintainability across the codebase.
|
|
Developed three distinct implementations of the Battleships game to explore multi-paradigmatic problem solving. Contrasted Rust's functional safety and immutability with Prolog's declarative logic resolution and Python's Object-Oriented state management. Analyzed execution flow and performance trade-offs across paradigms.
|
|
Responsive front-end development projects emphasizing modern UI/UX principles. Features semantic HTML5 architecture, adaptive CSS layouts using Flexbox and Grid, and JavaScript interactivity to ensure seamless cross-platform accessibility.
|
