Projects
Blind Backrunning With Fully Homomorphic Encryption
This protocol utilises Fully Homomorphic Encryption (FHE) to enable searchers to blindly backrun private user transactions on decentralised exchanges on the Ethereum blockchain.
Brain Graph Prediction Using Generative GNN's
This machine learning model leverages the capabilities of generative Graph Neural Networks (GNNs) to predict high-resolution brain MRIs from their low-resolution counterparts.
Neural Network for House Price Prediction
This project implements a custom, low-level implementation of a multi-layered neural network, built from scratch and incorporating a backpropagation algorithm to predict house prices in California. This model was developed entirely from scratch, without relying on any external machine learning libraries.
Decision Tree for Location Detection Using Wi-Fi Signal Strengths
A decision tree algorithm designed to accurately determine indoor locations by analysing Wi-Fi signal strengths collected from a mobile phone. This approach leverages the unique signal profiles to provide precise location estimations within indoor environments.
On-Chain Ticketing Service using NFTs
This on-chain ticketing service consists of three core components: a non-fungible token (NFT) contract that defines the ticket logic, a primary marketplace where users can create tickets by deploying new instances of the NFT contract, and a secondary marketplace that allows users to sell and place bids for re-sold tickets.
PCAP Visualiser for Improved Network Traffic Analysis
A packet visualiser which enhances the traditional Wireshark user interface by providing advanced filtering and grouping capabilities, enabling users to conduct more efficient analysis of network traffic.