Big Data Analysis of Accounting Forecasting Based on Machine Learning

Artificial intelligence, big data and other technologies will bring transformative significance to the traditional accounting and audit process, but the relevant research on the specific application methods of such technologies in the field of accounting and audit is still basically blank. In this paper, we combine 5 million pieces of user behavior information of Ant Financials’ Yu Ebao business to predict the company’s total daily capital inflow as the goal, and discuss the value of machine learning method in artificial intelligence in accounting prediction and specific implementation methods. In addition, based on the construction of application examples and the comparison between machine learning method and traditional method for accounting prediction, this paper also provides theoretical support for the application of machine learning method in the field of accounting. The results of this study show that machine learning is a very special method to predict the problem compared with traditional accounting, because it transcends the existing accounting information model to some extent. Although there has been a long history of using data to support accounting prediction, the depth, breadth, cost performance and potential of accounting prediction through artificial intelligence machine learning are unprecedented. The fundamental goal of machine learning technologies is to help a wide variety of organizations use modern data sets to gain valuable insights, and this is partly in line with the role that accounting has played in business operations in the past.