This repository demonstrates work completed as part of the CS5014 Machine Learning module for Practical 1. A machine learning model has been created that can predict credit card approvals from the described dataset using Logistic Regression. There are two deliverables: the source code for the solution, and a brief report which answers specific questions about the solution.
The aims of this practical were to gain experience in applying machine learning methodology to a real dataset. The submission demonstrates an understanding of:
- How to load, clean, and process a dataset;
- How to train a standard algorithm;
- How to report and interpret the results; and
- How to write clear, concise, and re-usable research code.
The dataset contains two files: crx.names contains the description of the different features in the dataset, while crx.data contains the actual data, in the CSV format. It contains anonymised data pertaining to individual credit card applications, as well as the final outcome (successful or unsuccessful). Notice that some of the values are numerical (also referred to as ‘continuous’), and some are categorical. The target variable (outcome of the application) uses ‘+’ for successful and ‘-’ for unsuccessful applications.