EFFECT OF KERNEL FUNCTION VARIATIONS ON AGRICULTURAL INDUSTRY SECTOR STOCK PREDICTION USING SUPPORT VECTOR REGRESSION METHOD

• Main Author: Rahmah, Gesti
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/49349
• Institution: Institut Teknologi Bandung

Econophysics is a branch of physics that applies concepts in physics to solve economic problems. One of the interesting things is stochastic dynamics in stock data. Stock data is included in stochastic dynamics because it has a pattern that changes with time that tends to fluctuate and is difficult to predict. In this study, the physical economy method support vector regression will be used to predict the closing stock data of the agricultural sector in eight Indonesian companies. Agriculture is a sector that has a significant role for the Indonesian economy. According to data from the Central Statistics Agency, in the second quarter of 2018 the contribution of agriculture to the growth rate of gross domestic product (GDP) reached 13.63%. The agriculture sector is very promising to be a business and a business and investment destination. Based on these conditions, it is necessary to have a model to predict the condition of the stock price index in the agricultural industry sector to assist investors in making decisions in investment or stocks. Prefrocessing data will be performed before it is used as data in the SVR learning process. Preprocessing data divides data into two parts, 80% of training data and 20% of test data. After obtaining the prediction results for the stock price index, an analysis of accuracy is performed by calculating the error value and comparing the effect of variations Kernel function on the accuracy of predictions. From the simulation results it was found that the agricultural stock sector data have almost the same characteristics. This is indicated by the use of a polynomial Kernel Function that produces a minimum error value in 50% of the number of companies with the smallest average error, namely MAPE = 13.10% and NRMSE = 0.07%.