| Autor: |
Abdulrauph Olarewaju, BABATUNDE, Uchenna Harrison, IGBOELI, Ghaniyyat Bolanle, BALOGUN, Idowu Dauda, OLADIPO, Gbemisola, BABALOLA, Gbolahan, OLASINA, Adeyinka, TELLA |
| Předmět: |
|
| Zdroj: |
Journal of Computer Science & Control Systems; May2020, Vol. 13 Issue 1, p17-24, 8p |
| Abstrakt: |
Information guides and strategies for ensuring consistent profiting from the stock market abound almost everywhere from magazines to professional journals and the internet. This mass volume of information sometimes confuses the investor who finds it difficult to sift the right strategy to use to be able to make a meaningful decision. This calls for the role of an artificially intelligent system that can use its knowledge base, proven algorithm coupled with an interactive and user-friendly interface to make an efficient and informed decision. In this paper, the researchers developed a framework for the prediction of stock prices for stocks listed in the Nigeria Stock Exchange. The research focuses on the applicability of an artificial intelligence-based model in the prediction of stock prices in the Nigeria market. Artificial Neural Network (ANN), Autoregressive Integrated Moving Average and (ARIMA) and Generalized Autoregressive Conditional Heteroskedasticity (GARCH) models were considered as individual agents in the experiment. Mean Absolute Error (MAE) Mean Square Error (MSE) and Mean Absolute Percentage Error (MAPE) evaluation metrics were used to evaluate the performance of the system with ten stocks from the Nigeria Stock Exchange. The results obtained prove that any of these methods can be used to predict stocks with over 90% accuracy level. An integrated system which promises to have a better prediction value is proposed. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Supplemental Index |
| Externí odkaz: |
|
