Quantitative trading system based on machine learning in Chinese financial market

The chapter describes the characteristics of machine learning methods in their possible application in investment portfolio optimization. With the use of the SWOT analysis, the features of the algorithms responsible for their increasing popularization in the formulation of investment strategies and their limitations in this regard were discussed. The prospects for further development of machine learning were described in the context of the market and technological environment. In addition, based on the review of the research, the possibilities of using machine learning algorithms in managing the investment portfolio and the use of modern research methods, which can be a creative development of the needs and solution to the problems faced by researchers of financial science and financial market practitioners, have been presented.

Download Full-text

Application of Machine Learning Technology in Financial Market

2020 International Conference on Big Data Application & Economic Management (ICBDEM 2020) ◽

10.25236/icbdem.2020.013 ◽

2020 ◽

Keyword(s):

Machine Learning ◽

Financial Market ◽

Learning Technology

Download Full-text

Forward Forecast of Stock Price Using LSTM Machine Learning Algorithm

International Journal of Computer Theory and Engineering ◽

10.7763/ijcte.2020.v12.1267 ◽

2020 ◽

Vol 12 (3) ◽

pp. 74-79

Author(s):

Kavitha Esther Rajakumari ◽

◽

M. Srinivasa Kalyan ◽

M. Vijay Bhaskar

Keyword(s):

Machine Learning ◽

Stock Price ◽

Short Term Memory ◽

Learning Algorithm ◽

Neural System ◽

Short Term ◽

Trading System ◽

Term Memory ◽

The Future ◽

Long Short Term Memory

Stock market prediction is the demonstration of attempting to decide the future estimation of an organization stock or other monetary instrument exchanged on a trade. This paper will exhibit how to perform stock expectations utilizing Machine Learning calculations. Foreseeing securities exchange costs is an intricate assignment that generally includes broad human-PC communication. Because of the connected idea of stock costs, customary bunch preparing techniques can't be used productively for securities exchange examination. In the current framework, the Sliding window calculation is used. This calculation investigates the information, with a window pushing ahead, in the wake of examining the information. It is very tedious for expectation of stocks. While, in the proposed framework, the utilization of LSTM (Long Short Term Memory) calculation, gives compelling outcomes. While analyzing, the superfluous information is overlooked. The current framework is additionally not viable, in taking care of non-straight information. What's more, it is less proficient contrasted with LSTM algorithm. So, to help defeat these, LSTM helps in dealing with the information in a productive way. Indeed, speculators are exceptionally intrigued by the exploration zone of stock value expectations. For decent and fruitful speculation, numerous financial specialists are sharp in knowing the future circumstance of the share trading system. Great and viable expectation frameworks for securities exchange encourage brokers, financial specialists, and investigators by giving steady data like the future course of the share trading system. In this work, an intermittent neural system (RNN) and Long Short-Term Memory (LSTM) are presented, a way to deal with anticipate securities exchange lists. The proposed model is a promising prescient procedure for a very non-direct time arrangement, whose designs are hard to catch by customary models.

Download Full-text

Blockchain Equity System Transaction Method and System Research Based on Machine Learning and Big Data Algorithm

Wireless Communications and Mobile Computing ◽

10.1155/2021/3457967 ◽

2021 ◽

Vol 2021 ◽

pp. 1-14

Author(s):

Kanghua Peng

Keyword(s):

Machine Learning ◽

Big Data ◽

Operating Efficiency ◽

Analysis Method ◽

Trading System ◽

Single Node ◽

Blockchain Technology ◽

Equity Trading ◽

Method Factor ◽

Depth Study

With the development of machine learning and big data, traditional equity trading system methods can no longer meet the current trading needs, and there are still problems such as low operating efficiency and serious homogeneity. Blockchain technology has the characteristics of decentralization and can also complete transactions through smart contracts, innovating the way of equity system transactions. The purpose of this paper is to build an equity trading system in combination with blockchain in the context of machine learning and big data and provide innovative trading methods, so as to provide reference and reference significance for the construction of my country’s equity market. This article uses literature data method, comparative analysis method, factor analysis method, and other methods to carry out research, in-depth study of machine learning and big data, blockchain-related concepts, system composition, application situation, etc., and discusses the allocation of equity trading market The functions of resources, risk diversification, risk transfer, price determination, etc., have built a blockchain equity trading system, designed a consensus mechanism, block generation protocol, block verification, decentralization, and smart contract platform, and finally conducted a national equity transaction the background of the market is analyzed, and the experimental results, simulation indicators, transaction time, transmission consumption, and other content of the system constructed in this article are analyzed. In the single-node test, the CPU usage of the PoW consensus mechanism algorithm reached 100%, but the improved PBFT consensus mechanism was only 16%, which saved a lot of computing power and improved computing performance.

Download Full-text

Modelling of stock market security price Dynamics Using market microstructure Data

Finance Theory and Practice ◽

10.26794/2587-5671-2018-22-5-141-153 ◽

2018 ◽

Vol 22 (5) ◽

pp. 141-153

Author(s):

N. A.  Bilev

Keyword(s):

Neural Network ◽

Machine Learning ◽

Logistic Regression ◽

Stock Market ◽

Regression Model ◽

Market Microstructure ◽

Logistic Regression Model ◽

Network Models ◽

Neural Network Models ◽

Trading System

In modern electronic stock exchanges there is an opportunity to analyze event driven market microstructure data. This data is highly informative and describes physical price formation which makes it possible to find complex patterns in price dynamics. It is very time consuming and hard to find this kind of patterns by handcrafted rules. However, modern machine learning models are able to solve such issues automatically by learning price behavior which is always changing. The present study presents profitable trading system based on a machine learning model and market microstructure data. Data for the research was collected from Moscow stock exchange MICEX and represents a limit order book change log and all market trades of a liquid security for a certain period. Logistic regression model was used and compared to neural network models with different configuration. According to the study results logistic regression model has almost the same prediction quality as neural network models have but also has a high speed of response which is very important for stock market trading. The developed trading system has medium frequency of deals submission that lets it to avoid expensive infrastructure which is usually needed in high-frequency trading systems. At the same time, the system uses the potential of high quality market microstructure data to the full extent. This paper describes the entire process of trading system development including feature engineering, models behavior comparison and creation of trading strategy with testing on historical data.

Download Full-text