What is it about?

Stock is a multiplayer game, and its price fluctuations are often affected by a variety of complex factors: economic cycles, changes in the country's financial situation; national policy adjustments or changes; the issuing company's operating performance, and other factors of the company itself; changes in the industry's position in the national economy; investor movements, the intentions and manipulations of large investors; and the psychological state of investors after being influenced by various aspects of change. Some factors are difficult to measure and record, such as human investment psychology or the manipulation intentions of large investors. Some factors are difficult to quantify, so it is extremely difficult to try to predict changes in stock prices by these factors. When the causal mapping process is ignored, the difficulty of prediction can be reduced by directly studying the results of the combination of these factors and exploring the patterns between the different results. Currently, the use of long and short-term memory neural networks in stock price prediction is a feasible approach that is commonly accepted.

Featured Image

Perspectives

So far, some are also scholars who use long and short-term memory neural networks to do stock price prediction, but theirs are more about stock-specific and do not give a solution when dealing with a large number of stocks. When dealing with a large number of different stocks, their solutions require building models for each stock. This paper achieves a fast prediction of individual stock prices of the A-share Shanghai exchange by using the K-means unsupervised clustering method and long and short-term memory neural network.

Xingyu He

Read the Original

This page is a summary of: Prediction of A-share Prices Based on K-means Clustering Algorithm and LSTM, August 2022, ACM (Association for Computing Machinery),
DOI: 10.1145/3561801.3561804.
You can read the full text:

Read

Contributors

The following have contributed to this page