Indicator selection with committee decision of filter methods for stock market price trend in ISE

Pehlivanl, Aya akmak; Akgil, Bar; Glay, Gzhan

doi:10.1016/j.asoc.2016.09.004

Cited by 22 publications

(6 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent studies, it has been observed that accuracy is often used as an evaluation metric [35]. Along with the accuracy metric, there are also studies using sensitivity, specificity, and area under the curve metrics [9,19]. Accuracy can be misleading if the datasets have an imbalanced class distribution.…”

Section: Discussionmentioning

confidence: 99%

“…• As the use of other stock features in the classification extends our feature space, dimensionality reduction is required. In recent prediction studies, dimensionality reduction is performed by a filter approach that evaluates the relation between each feature and class labels [9,19]. In our study, as the number of dimensions increased up to 5660, filter-based and computationally efficient methods, relief and gain ratio, were chosen.…”

Section: Discussionmentioning

confidence: 99%

“…To the best of our knowledge, this is the first study that uses features computed on all other stocks to predict the movement of a particular stock in BIST. Although previously individual stock features have been used for the prediction of a stock in numerous studies [9,10] and correlations between stock prices have been investigated [11], to the best of our knowledge, features of other stocks have not yet been used for prediction in the Turkish markets. Our second contribution is the use of different types of feature selection and classification methods together.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Stock daily return prediction using expanded features and feature selection

Gündüz¹,

Çataltepe²,

Yaslan³

2017

Turk J Elec Eng & Comp Sci

View full text Add to dashboard Cite

Stock market prediction is a very noisy problem and the use of any additional information to increase accuracy is necessary. In this paper, for the stock daily return prediction problem, the set of features is expanded to include indicators not only for the stock to be predicted itself but also a set of other stocks and currencies. Afterwards, different feature selection and classification methods are utilized for prediction. The daily close returns of the 3 most traded stocks (GARAN, THYAO, and ISCTR) in Borsaİstanbul (BIST) are predicted using indicators computed on those stocks, indicators for all the other stocks listed in the BIST100 index, and indicators on the dollar-gold prices. Twenty-five different indicators on daily stock prices are computed to form feature vectors for each trading day. These feature vectors are assigned class labels according to the daily close returns. Expanding the feature space with BIST100 stocks features results in a high dimensional feature space, with possibly noisy or irrelevant features. Therefore, feature selection methods are utilized to select the most informative features. In order to determine relevance scores of features, fast filter-based methods, gain ratio and relief, are used. Experiments are performed based on individual stock features, dollar-gold features (DG), BIST100 stock features (BIST100), and a combination of BIST100 and DG with and without feature selection. Using the gain ratio feature selection with a gradient boosting machine (GBM), the movements of GARAN stock were predicted with an accuracy of 0.599 and an F-measure of 0.614. For THYAO, the relief feature selection with the GBM gave an accuracy of 0.558, and for ISCTR, the gain ratio feature selection with logistic regression achieved an accuracy of 0.581. It was found that using BIST100 stock features boosts classification results for all stocks in terms of accuracy.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Stock daily return prediction using expanded features and feature selection

Gündüz¹,

Çataltepe²,

Yaslan³

2017

Turk J Elec Eng & Comp Sci

View full text Add to dashboard Cite

show abstract

“…It is shown that NN gives better results. In [41], optimal subset indicators are selected with ensemble feature selection approach in order to increase the performance of predicting the next day's stock price direction. A real dataset is obtained from Istanbul Stock Exchange (ISE), and the subset is composed using technical and macroeconomic indicators.…”

Section: Related Workmentioning

confidence: 99%

Modeling Traders’ Behavior with Deep Learning and Machine Learning Methods: Evidence from BIST 100 Index

2020

View full text Add to dashboard Cite

Although the vast majority of fundamental analysts believe that technical analysts’ estimates and technical indicators used in these analyses are unresponsive, recent research has revealed that both professionals and individual traders are using technical indicators. A correct estimate of the direction of the financial market is a very challenging activity, primarily due to the nonlinear nature of the financial time series. Deep learning and machine learning methods on the other hand have achieved very successful results in many different areas where human beings are challenged. In this study, technical indicators were integrated into the methods of deep learning and machine learning, and the behavior of the traders was modeled in order to increase the accuracy of forecasting of the financial market direction. A set of technical indicators has been examined based on their application in technical analysis as input features to predict the oncoming (one-period-ahead) direction of Istanbul Stock Exchange (BIST100) national index. To predict the direction of the index, Deep Neural Network (DNN), Support Vector Machine (SVM), Random Forest (RF), and Logistic Regression (LR) classification techniques are used. The performance of these models is evaluated on the basis of various performance metrics such as confusion matrix, compound return, and max drawdown.

show abstract

“…A genetic algorithm-based approach to feature discretization in artificial neural network has been used for the forecast of the stock price index (Kim & Han, 2000). In (Pehlivanli, Asikgil & Gulay, 2016) the support vector machines (SVMs) are combined with four filter methods based on different metrics to obtain filtered features for forecasting stock prices for Istanbul Stock Exchange market. A hybrid model that consists of two linear models (autoregressive moving average model, exponential smoothing model) and a recurrent neural network were used for the prediction of stock returns (Rather, Agarwal & Sastry, 2015).…”

Section: Introductionmentioning

confidence: 99%

Competitive / Collaborative Statistical Learning Framework for Forecasting Intraday Stock Market Prices: A Case Study

Belciug¹,

Săndiță²,

Costin³

et al. 2021

STUD INFORM CONTROL

View full text Add to dashboard Cite

This paper presents an intelligent decision system based on statistical learning that regards the tactics of an investor in predicting the next intraday stock price. Significant percentages can be won or lost depending on the tactics applied for buying/selling shares. This paper includes a case study regarding the efficiency of a group of machine learning techniques that work together in a competitive/collaborative manner with a view to achieving an overall price forecast for the next intraday transaction. In order to illustrate the advantages of this intelligent decision system this work provides a concrete example concerning the price forecast for the next intraday transaction for Transilvania Bank (TLV), the stock market at the Bucharest Stock Exchange (BVB), Romania. An important part of the decision system lies in the competitive stage, because only the best competitors are chosen for the ultimate decision-making process. In the collaborative stage of the statistical learning framework one uses a weighted voting system that outputs the final intraday stock price. The results obtained show that this intelligent system outperforms each stand-alone method.

show abstract

Indicator selection with committee decision of filter methods for stock market price trend in ISE

Cited by 22 publications

References 27 publications

Stock daily return prediction using expanded features and feature selection

Stock daily return prediction using expanded features and feature selection

Modeling Traders’ Behavior with Deep Learning and Machine Learning Methods: Evidence from BIST 100 Index

Competitive / Collaborative Statistical Learning Framework for Forecasting Intraday Stock Market Prices: A Case Study

Contact Info

Product

Resources

About