Attention based adaptive spatial-temporal hypergraph convolutional networks for stock price trend prediction

Hongyang Su, Xiaolong Wang, Yang Qin, Qingcai Chen

Abstract: Highlights•An adaptive framework for stock price trend prediction without any prior information.•The invalid correlations of stocks are identified and filtered.•Extensive evaluations show accurate forecasting of trends and minor fluctuations.•Several practical applications of simulated investments are demonstrated.•Interpretability studies of the model describe the valid stock correlations.