| Abstract |
Fake accounting records result from accounting fraud which shows excess income, profits, revenues, assets, and sales, or underreports costs, debts, losses, and expenses. To analyse such fraudulent accounting statements, conventional methods like audit and physical review are time-consuming, expensive, and inaccurate. Smart methods can greatly assist accountants in studying numerous accounting records in a very less amount of time. Machine learning (ML) can be leveraged in financial management within smart cities to enhance efficiency, accuracy, and decision-making processes. An overview of different ML and deep learning (DL) strategies used to identify financial fraud in smart cities is presented in this chapter. This review focuses on ML and DL technologies for identifying fraud in financial statements. The major difficulties, research gaps, and challenges in the research area of financial statement fraud detection are discussed along with future scope. Previous studies utilized trained or supervised algorithms more often than unpredictive or unsupervised algorithms such as clustering, so future researchers in fraud detection can obtain better results by using advanced learning mechanisms like deep learning. © 2024 selection and editorial matter, Om Prakash Mahela, Baseem Khan and Puneet Kumar Jain. individual chapters, the contributors. |
