Construction of Quantitative Transaction Strategy Based on LASSO and Neural Network

Xu Wang, Jia-Yu Zhong and Zi-Yu Li

Applied Economics and Finance, 2017, vol. 4, issue 4, 134-144

Abstract: Since the establishment of the securities market, there has been a continuous search for the prediction of stock price trend. Based on the forecasting characteristics of stock index futures, this paper combines the variable selection in the statistical field and the machine learning to construct an effective quantitative trading strategy. Firstly, the LASSO algorithm is used to filter a large number of technical indexes to obtain reasonable and effective technical indicators. Then, the indicators are used as input variables, and the average expected return rate is predicted by neural network. Finally, based on the forecasting results, a reasonable quantitative trading strategy is constructed. We take the CSI 300 stock index futures trading data for empirical research. The results show that the variables selected by LASSO are effective and the introduction of LASSO can improve the generalization ability of neural network. At the same time, the quantitative trading strategy based on LASSO algorithm and neural network can achieve good effect and robustness at different times.

Keywords: quantitative transaction strategy; LASSO; neural network (search for similar items in EconPapers)
Date: 2017
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://redfame.com/journal/index.php/aef/article/view/2370/2653 (application/pdf)
http://redfame.com/journal/index.php/aef/article/view/2370 (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:rfa:aefjnl:v:4:y:2017:i:4:p:134-144

Access Statistics for this article

More articles in Applied Economics and Finance from Redfame publishing Contact information at EDIRC.
Bibliographic data for series maintained by Redfame publishing ().