Stock Price Prediction Based on Technical Indicators with Soft Computing Models
Stock market prediction is a very tough task in the finance world. Since stock prices are dynamic, noisy, non-scalable, non-linear, non-parametric and complicated. In recent years, soft computing techniques are used for developing stock prediction model.
- PDF / 2,260,192 Bytes
- 15 Pages / 439.37 x 666.142 pts Page_size
- 48 Downloads / 210 Views
School of Business and Management, CHRIST (Deemed to Be University), Bangalore, India [email protected] 2 Department of Management Studies, Christ University, Bangalore, India
Abstract. Stock market prediction is a very tough task in the finance world. Since stock prices are dynamic, noisy, non-scalable, non-linear, non-parametric and complicated. In recent years, soft computing techniques are used for developing stock prediction model. The main focus of this study is to develop and compare the efficiency of the three different soft computing techniques for predicting the intraday price of individual stocks. The proposed models are based on Time Delay Neural Network (TDNN), Radial Basis Function Neural Network (RBFNN) and Back Propagation Neural Network (BPNN). The predictive models are developed using technical indicators. Sixteen technical indicators were calculated from the historical price and used as inputs of the developed models. Historical prices from 01/01/2018 to 28/02/2018, where the time interval between samples is one minute, are utilized for developing models. The performance of the proposed models is evaluated by measuring some metrics. Also, this study compares the results with other existing models. The experimental result revealed that the BPNN outperforms TDNN, RBFNN as well as other existing models considered for comparison. Keywords: Back Propagation Neural Network (BPNN) Intraday stock price prediction MAPE MAE Technical indicators Time Delay Neural Network (TDNN)
1 Introduction Prediction of the stock price is one of the most important issues in the financial world that has received more attention from financial analysts and researchers [1]. Stock market prediction is defined as the process of calculating the future value of stockbased on historical prices. It is regarded as a challenging task due to the nature of stock prices such as highly noisy, complicated, dynamic, irregular, nonlinear, chaotic and non-parametric [2–4]. Further to this, the financial market is also influenced by many external factors including economic conditions, political events and traders’ expectations etc. This fact motivates the researchers to develop a more efficient predictive model to make a good decision and gain more profits. Several traditional methods can be used to predict the stock market price which includes fundamental analysis, statistical analysis and technical analysis [5, 6]. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 J. I.-Z. Chen et al. (Eds.): ICIPCN 2020, AISC 1200, pp. 685–699, 2021. https://doi.org/10.1007/978-3-030-51859-2_62
686
S. Kumar Chandar
Fundamental analysis uses macro-economic factors, financial conditions, exchange rate and other factors to determine stock price. On the contrary, technical analysis involves historical prices. The technical analysis determines the stock value based on low, high, open and closing value. However, these traditional techniques cannot predict the stock market pri
Data Loading...
