Using Machine Learning to Predict Personal Expenditure
A complete understanding of personal finances is becoming increasingly important as the average persons disposable income has decreased due to a changing financial climate.
The aim of this project is to build an application that makes it easier to manage a users personal finances. This is split into two halves, accessing historical information in an easy to understand way and using machine learning techniques to predict future financial transactions. The security considerations of storing personal finance information are also considered.
This begins with a review of the existing commercial personal finance applications and the current techniques used to forecast time-boxed financial data, such as the value of a stock on the stock market, before detailing the design and implementation of the application.
Having completed the application, the performance of selected techniques is reviewed, before discussing further research opportunities which could improve the applications accuracy.
Project Title: Using Machine Learning to Predict Personal Expenditure
Author: Pez Cuckow
Degree: Computer Science with Business and Management
Supervisor: Dr Gavin Brown
Keywords: Markov Chain Models, Weighted Arithmetic Mean, Responsive Web Design, Web System Security
Content, ideas and output all copyright Pez Cuckow.