DOI: 10.5121/ijci.2023.120201

Improving CNN-base Stock Trading By Considering Data Heterogeneity and Burst

Keer YangGuanqun ZhangChuan BiQiang GuanHailu XuShuai Xu

Mar 2023

摘要

In recent years, there have been quite a few attempts to apply intelligenttechniques to financial trading, i.e., constructing automatic and intelligenttrading framework based on historical stock price. Due to the unpredictable,uncertainty and volatile nature of financial market, researchers have alsoresorted to deep learning to construct the intelligent trading framework. Inthis paper, we propose to use CNN as the core functionality of such framework,because it is able to learn the spatial dependency (i.e., between rows andcolumns) of the input data. However, different with existing deeplearning-based trading frameworks, we develop novel normalization process toprepare the stock data. In particular, we first empirically observe that thestock data is intrinsically heterogeneous and bursty, and then validate theheterogeneity and burst nature of stock data from a statistical perspective.Next, we design the data normalization method in a way such that the dataheterogeneity is preserved and bursty events are suppressed. We verify outdeveloped CNN-based trading framework plus our new normalization method on 29stocks. Experiment results show that our approach can outperform othercomparing approaches.

展开全部

图表提取

论文十问

参考文献

被引用

提问与回答

暂无人提供速读十问回答

被引用

笔记

问答

领域