Machine Learning • Optimisation • Data Analytics
I am a recent Data Analytics & Physics graduate looking to begin a career within data science and software engineering. Below you will find some of the recent projects on which I have worked along with all relevant information and contact links.
A dissertation project exploring modern optimisation algorithms aimed at accelerating gradient descent used in training machine learning models. Motivated by the lack of a comprehensive comparison in literature, the project looked to implement and compare these algorithms in different machine learning problems.
A sudoku solver designed to solve a sudoku from an input image. The project involves a data pipeline for extracting the sudoku grid from the image, this is then passed to the CNN which identifies each digit within the grid. Finally, the scanned grid is fed into a backtracking algorithm producing a solved grid.
A C++ implementation of a neural network using only the standard libraries. Comprises of neuron and network, capable of constructing arbitrary topologies with gradient descent training using backpropagation.
Python-based client and server used for setting up a chatroom over a local network. Utilises the socket module in Python to establish a connection handshake with the server.
Analysis of time series data sourced from AirBnB. Used to investigate and visualise the impact of the Covid19 pandemic on the hospitality industry around the globe.
A JavaScript project aimed at producing an interactive visualisation of the training process involving a machine learning model.
A project aimed at utilising the functionality of the D3.js library to produce an interactive visualisation of player data sourced from Fifa 20.
PyTorch object tracking algorithm compatible with any arbitrary detector capable of producing an object detection along with the bounding box and confidence scores.