6006CEM Machine Learning and Related Applications (2223SEPJAN)

This repository contains the implementation and demonstration files for 6006CEM, delivered in semester 1 starting September 2022 and to be completed by December 2022.

Chosen machine learning dataset: Rain In Australia - https://www.kaggle.com/datasets/jsphyg/weather-dataset-rattle-package

Directory Structure and Instructions

This repository contains the following files and folders:

• data/: This directory contains the dataset weatherAUS.csv.
• tuning/: A folder containing tuning logs for XGB and SVM models.
• figs/: A directory which holds some correlation figures (only the ones in the report).
• code.ipynb: This file contains the main notebook used to implement the machine learning algorithms.
• code_no_markdown.py: A python script with all the code and markdown cells commented out. Note: never been run! Only created to paste into the appendix section of the report.
• requirements.txt: A list of dependencies for this project to be installed via pip.
• module_check.ipynb: An interactive python notebook used to check if the required modules are installed.

To install all the requirements automatically, run the command pip install -r requirements.txt. You can use the supplied module_check.ipynb notebook to verify the requirements have been met.

Python and Module Versions

This code is tested to be working with the following modules and versions:

• Python version: 3.10.8 | packaged by conda-forge | (main, Nov 24 2022, 14:07:00) [MSC v.1916 64 bit (AMD64)]
• pandas version: 1.5.2
• matplotlib version: 3.6.2
• NumPy version: 1.23.5
• SciPy version: 1.9.3
• IPython version: 8.7.0 (part of jupyter notebook / jupyterlab)
• scikit-learn version: 1.1.3
• seaborn version: 0.12.1
• XGBoost version: 1.7.1