Final project for BSc Computer Science at the University of Exeter
This project looks into the machine learning and optimisation methods used to find gene-gene interaction in a Genome-Wide Association Study(GWAS). These association are found by measuring differences in genetic variance(Single Nucleotide Polymorphisms) between cases and controls, in a GWAS. The project consists of three development chunks, the data representation, the objective functions and the algorithms. This project looks at two stochastic search algorithms, Ant-Colony Optimisation and Ant-Colony Optimisation with Tabu List, and one deep learning generative causal model, The Implicit Causal Model. The paper mainly focuses on finding two locus analyses in the Type-II diabetes dataset.