Robust stock market mining, incorporating financial news and technical analysis data

Manolis Maragoudakis, Dimitrios Serpanos

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Stock market prediction has always gained certain attention from researchers, hence the plethora of analysis methods and tools. The majority of them are strongly related to structured, numerical databases and domain expertise rules. In the field of trading, most of decision support tools focus on statistical analysis of past price records. Nevertheless, throughout recent studies, prediction is also based on textual data, based on the rational assumption that the course of a stock price can be influenced by news articles, ranging from companies releases and local politics to news of superpower economy. The present paper deals with a special Random Forest Data mining technique, designed to alleviate the significant issue of high dimensionality in volatile and complex domains, such as stock market prediction. Information from both technical analysis as well as textual data from various on-line financial news resources is effectively parsed by our proposed hybrid ensemble algorithm, named as MBRF. Different experiments are carried out to evaluate different aspects of the problem, returning satisfactory results. The results show that the trading strategies guided by the proposed approach generate higher profits than the buy-and-hold strategy, as well as those guided by the level-estimation based forecasts of standard linear regression models and other Machine Learning classifiers such as Support Vector Machines, traditional Random Forests or Neural Networks.

Original languageEnglish
Pages (from-to)179-191
Number of pages13
JournalEngineering Intelligent Systems
Volume18
Issue number3-4
Publication statusPublished - 1 Sep 2010
Externally publishedYes

Fingerprint

stock market
prediction
superpower
data mining
Linear regression
Support vector machines
Data mining
Learning systems
Statistical methods
Profitability
statistical analysis
politics
Classifiers
Neural networks
resource
data analysis
Financial markets
Industry
experiment
Experiments

Keywords

  • Data mining
  • Expert systems
  • Markov blanket
  • Random Forests
  • Stock return forecasting
  • Trading strategies

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Control and Systems Engineering
  • Computer Science Applications

Cite this

Robust stock market mining, incorporating financial news and technical analysis data. / Maragoudakis, Manolis; Serpanos, Dimitrios.

In: Engineering Intelligent Systems, Vol. 18, No. 3-4, 01.09.2010, p. 179-191.

Research output: Contribution to journalArticle

Maragoudakis, Manolis ; Serpanos, Dimitrios. / Robust stock market mining, incorporating financial news and technical analysis data. In: Engineering Intelligent Systems. 2010 ; Vol. 18, No. 3-4. pp. 179-191.
@article{9a1739736ecd4f01b5214dea86dd96ac,
title = "Robust stock market mining, incorporating financial news and technical analysis data",
abstract = "Stock market prediction has always gained certain attention from researchers, hence the plethora of analysis methods and tools. The majority of them are strongly related to structured, numerical databases and domain expertise rules. In the field of trading, most of decision support tools focus on statistical analysis of past price records. Nevertheless, throughout recent studies, prediction is also based on textual data, based on the rational assumption that the course of a stock price can be influenced by news articles, ranging from companies releases and local politics to news of superpower economy. The present paper deals with a special Random Forest Data mining technique, designed to alleviate the significant issue of high dimensionality in volatile and complex domains, such as stock market prediction. Information from both technical analysis as well as textual data from various on-line financial news resources is effectively parsed by our proposed hybrid ensemble algorithm, named as MBRF. Different experiments are carried out to evaluate different aspects of the problem, returning satisfactory results. The results show that the trading strategies guided by the proposed approach generate higher profits than the buy-and-hold strategy, as well as those guided by the level-estimation based forecasts of standard linear regression models and other Machine Learning classifiers such as Support Vector Machines, traditional Random Forests or Neural Networks.",
keywords = "Data mining, Expert systems, Markov blanket, Random Forests, Stock return forecasting, Trading strategies",
author = "Manolis Maragoudakis and Dimitrios Serpanos",
year = "2010",
month = "9",
day = "1",
language = "English",
volume = "18",
pages = "179--191",
journal = "JAPCA",
issn = "1073-161X",
publisher = "Taylor and Francis Ltd.",
number = "3-4",

}

TY - JOUR

T1 - Robust stock market mining, incorporating financial news and technical analysis data

AU - Maragoudakis, Manolis

AU - Serpanos, Dimitrios

PY - 2010/9/1

Y1 - 2010/9/1

N2 - Stock market prediction has always gained certain attention from researchers, hence the plethora of analysis methods and tools. The majority of them are strongly related to structured, numerical databases and domain expertise rules. In the field of trading, most of decision support tools focus on statistical analysis of past price records. Nevertheless, throughout recent studies, prediction is also based on textual data, based on the rational assumption that the course of a stock price can be influenced by news articles, ranging from companies releases and local politics to news of superpower economy. The present paper deals with a special Random Forest Data mining technique, designed to alleviate the significant issue of high dimensionality in volatile and complex domains, such as stock market prediction. Information from both technical analysis as well as textual data from various on-line financial news resources is effectively parsed by our proposed hybrid ensemble algorithm, named as MBRF. Different experiments are carried out to evaluate different aspects of the problem, returning satisfactory results. The results show that the trading strategies guided by the proposed approach generate higher profits than the buy-and-hold strategy, as well as those guided by the level-estimation based forecasts of standard linear regression models and other Machine Learning classifiers such as Support Vector Machines, traditional Random Forests or Neural Networks.

AB - Stock market prediction has always gained certain attention from researchers, hence the plethora of analysis methods and tools. The majority of them are strongly related to structured, numerical databases and domain expertise rules. In the field of trading, most of decision support tools focus on statistical analysis of past price records. Nevertheless, throughout recent studies, prediction is also based on textual data, based on the rational assumption that the course of a stock price can be influenced by news articles, ranging from companies releases and local politics to news of superpower economy. The present paper deals with a special Random Forest Data mining technique, designed to alleviate the significant issue of high dimensionality in volatile and complex domains, such as stock market prediction. Information from both technical analysis as well as textual data from various on-line financial news resources is effectively parsed by our proposed hybrid ensemble algorithm, named as MBRF. Different experiments are carried out to evaluate different aspects of the problem, returning satisfactory results. The results show that the trading strategies guided by the proposed approach generate higher profits than the buy-and-hold strategy, as well as those guided by the level-estimation based forecasts of standard linear regression models and other Machine Learning classifiers such as Support Vector Machines, traditional Random Forests or Neural Networks.

KW - Data mining

KW - Expert systems

KW - Markov blanket

KW - Random Forests

KW - Stock return forecasting

KW - Trading strategies

UR - http://www.scopus.com/inward/record.url?scp=84866403098&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84866403098&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:84866403098

VL - 18

SP - 179

EP - 191

JO - JAPCA

JF - JAPCA

SN - 1073-161X

IS - 3-4

ER -