<![CDATA[One of the advantages of investing in stocks is capital gains, which is the benefits from stock trading. The use of stock price forecasting will increase profits from stock sale and purchase transactions because shareholders can know when the right time to sell or buy a particular stock. Support Vector Regression (SVR) is one of the methods used for forecasting, it can recognize patterns of time series data and can provide good forecasting results when the parameters of importance can be determined well as well. So we need an optimization method to determine SVR parameters so that SVR can be optimally applied in stock price forecasting. One of the optimization algorithms that can be used is Particle Swarm Optimization (PSO). Stock price forecasting using SVR with PSO optimization uses MAPE method to evaluate forecasting results. Based on the test that has been done, the value of MAPE obtained is 0.8195% with fitness of 0.5496 with the optimal parameters obtained is the number of particles 40, iteration PSO 40, iteration SVR 1000, parameter range of C 100 - 500, the parameter range of ɛ 0, 0001 - 0,001, parameter range of σ 0,001 - 2, parameter range of γ 0,00001 - 0,001, parameter range of λ 0,001 - 0,1, and comparison of training data and test data from 2016 BCA Bank share data is 90%: 10%.]]>
Copyrights © 2018
