This folder contains documentated codebases used to produce the results of the experiments presented in this thesis.
This folder is organized into seperate modules for each codebase:
.
├── cnn
├── data
├── ga
├── gp
├── pca
└── pso
With the exception of the data folder, which is used by all modules, each module is designed to function in isolation. Those modules are:
- cnn, an 1D convolutional neural network (CNN) for classification on Gas Chromatography data.
- data, this folder contains the Gas Chromatography data, stored as a matlab file. And the unprocessed raw data for archival purposes.
- ga, an implementation of a basic Genetic Algorithm (GA), that can solve basic 2nd order polynomials and onesum.
- gp, a Genetic Program (GP) for classification on the Fish Oil Data. It constructs a basic GP tree with arithmetic operators and a classification map (CM) to facilitate classification.
- julia, a Julia implementation of PSO for the michalewicz & mccormick combinatorial optimization problems.
- pca, Principal Component Analysis (PCA) implemented from scratch using Eigenvectors and Numpy library.
- pso, the feature selection codebase for the experiments for the AJCAI paper. This includes an implementation of Particle Swarm Optimisation.