Detecting Accounting Fraud in Publicly Traded Firms Using a Machine Learning Approach with Live Implementation

Abstract

In this research, the authors use a machine learning-based approach to creating a state-of-the-art fraud prediction model. It is beneficial to construct models using a combination of domain knowledge and machine learning techniques. The paper provides a literature review and classification for the investigation of data mining's function in identifying instances of financial fraud. In this paper, researchers used some techniques for detecting accounting fraud, i.e., Random Forest, Genetic Algorithm, Decision Tree, and Convolutional Neural Network (CNN). A comprehensive literature research on Financial Fraud Detection (FFD) has not been conducted despite the reality that it is a rapidly developing field of critical relevance. Financial accounting fraud has increased, making financial accounting fraud detection (FAFD) an issue of major relevance in academia, research, and industry due to the current financial environment. Forensic accounting techniques are now used to identify instances of financial accounting fraud due to the lack of effectiveness of the company's business auditing system. Data mining methods are proving to be of considerable assistance in the identification of fraud in financial accounting due to the difficulties of forensic accounting in handling high data quantities and the complexity of financial data. As a result, it can be used for a wide variety of things, including machine learning, web development, and software testing. Performing a visual representation of genuine data and fraudulent data, which are indicated by 0 and 1.