7/28/2019 Stock Prices Forecast Using Radial Basis Function Neural Network http://slidepdf.com/reader/full/stock-prices-forecast-using-radial-basis-function-neural-network 1/9 (IJCSIS) International Journal of Computer Science and Information Security, Vol.11, No.3, 2013 STOCK PRICES FORECAST USING RADIAL BASIS FUNCTION NEURAL NETWORK Julia Fajaryanti Faculty of Industrial Technology Gunadarma University Jalan Margonda Raya 100 Depok - Indonesia . Priyo Sarjono Wibowo Faculty of Industrial Technology Gunadarma University Jalan Margonda Raya 100 Depok - Indonesia Abstract — Neural Network has been implemented in various applications especially in pattern recognition. This power has attracted several people to use Neural Network for various systems. One of the neural network implementation in the field of finance or investments is forecasting stocks. Assuming that the prediction of the output system is deterministic, than the suitable Neural Network model to predict it is Multilayer Network. To get the solution, Multilayer Neural Network method with supervised algorithm is applied. The supervised algorithm used for stock price prediction is Radial Basis Function. This algorithm can supervise the networks by using previous stock price data, classifying them and putting weight on the networks. This journal illustrate how Radial Basis Function Neural Network method can be used to predict stocks. The result showed that Radial Basis Function Neural Network method is able to forecast and follow the movement of stock data used in the experiment. Keywords: Stock Prices, Multilayer Neural Network, Radial Basis Function, Supervised I. I NTRODUCTION The role of capital markets in the Indonesian economy began institutionalized. Currently one of the purchase of shares legitimate capital choices, in addition to other forms of capital such as money, land, and gold. Rational factors and various irrational factors be the deciding factor in purchasing shares. Rational factors commonly associated with the analysis fundamental. Fundamental analysis does not consider the pattern of movement of shares in the past but trying to determine the appropriate value for a stock. In perfect capital markets and efficient, stock prices reflect all publicly available information on and information exchange that can only be obtained from certain groups. High and low stock prices influenced by many factors such as conditions and company performance, risk, dividend, interest rates, conditions economy, government policy, inflation, supply and demand as well as many more. Because anticipate possible changes in the factors above, the stock price can rise or fall. Prediction of the stock price will very useful for investors to be able to see how the prospects of investing in the stock of a company come. Stock price prediction can be used to anticipate the rise and fall of stock prices. With the prediction of share price, it is very helpful for investors in decision making. There are two methods that prediction can use in this implementation, namely: conventional methods and Artificial Neural Network (ANN). In this journal, authors will implement Radial Basis Function Neural Network in financial application to forecast the stock prices. II. STATEOFTHEART There are many factors that influence the price of a stock. Most of these factors are included in the factors related to the social situation, where it is very difficult to estimate. However, there are things that can be used as a basis for estimating the price of a stock, one by studying the previous price of a stock, and then we can estimate the stock price in the future. This, of course can help decision-makers to the activities of buying and selling a stock [7]. For that we need a method to identify and study the movement of the stock price over time in order to estimate the price of a stock of a particular company. It also takes a special device to do it, this is due to the limitations of human beings in the data processing activities that are so difficult and numerous in numbers. Thus, we use the computer to do it. To be able to work on it, the computer requires a learning process. In this process, computers are faced with a series of data that has been classified and will study the patterns of the data. Lessons learned include the adjustment to a predetermined pattern, or by studying the similarity of the pattern. It is accompanied by the development of computers, the better, in terms of speed, accuracy, and cost [8]. Neural network is a system that can be relied to do a complex quick count processing with higher speed compared to another organs. The use of neural networks requires an understanding of the way of thinking in the human brain. So many elements in the human brain, where they are connected 21 http://sites.google.com/site/ijcsis/ ISSN 1947-5500
9
Embed
Stock Prices Forecast Using Radial Basis Function Neural Network
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
7/28/2019 Stock Prices Forecast Using Radial Basis Function Neural Network