A Comparison of SOM and K-Means Algorithms in Predicting Tax Compliance

View Sample PDF

Author(s): Felix Bankole (University of South Africa, South Africa)and Zama Vara (University of South Africa, South Africa)
Copyright: 2023
Pages: 21
Source title: Encyclopedia of Data Science and Machine Learning
Source Author(s)/Editor(s): John Wang (Montclair State University, USA)
DOI: 10.4018/978-1-7998-9220-5.ch155

Keywords: Data Analysis and Statistics / Engineering Science Reference / Knowledge Management and Library Science / Machine Learning

Purchase

View A Comparison of SOM and K-Means Algorithms in Predicting Tax Compliance on the publisher's website for pricing and purchasing information.

Abstract

Compliance with taxes has been associated with the perceived probability of detection and severity of punishment. Firstly, taxpayers comply because the benefits far outweigh the costs. Secondly, decisions to comply are closely related to the individual's risk aversion. Lastly, the availability of opportunities to cheat and the perceived probabilities of detection and sanctions have a significant impact on taxpayer compliance. In this study, two unsupervised learning algorithms were employed. The algorithms used were self-organizing map (SOM) and k-means clustering in predicting corporate income tax compliance. SOM and k-means are two notable unsupervised learning clustering techniques. In this article, some experiments have been conducted to compare their effectiveness and performance.

The IRMA Community

Research IRM

A Comparison of SOM and K-Means Algorithms in Predicting Tax Compliance

Purchase

Abstract

Related Content

IRMA Sponsors