Penerapan Least Squares Support Vector Machines (LSSVM) dalam Peramalan Indonesia Composite Index

In the era of very rapidly advancing technology like today, both internet and computerization have made various corporate agencies or investors start thinking about importance stock market in their capital division. Previously there were purchases by company's capital, such: gold, land, buildings, production machines, but at this time purchase shares should also to attract attention these are legal investments. Various kinds company that sold can already be seen through it is easy attractive for companies will make purchases, even model chosen long-term short-term purchases. This price forecasting system using Least Squares Support Vector Machines (LSSVM) method popular with help determine conclusions buying because reduce losses right decisions so increase profits companies. a simpler has been modified from previous model, namely: (SVM) method. Solving linear equations solved way LSSVM compared SVM. The variable used network close variable. kernel study RBF kernel. consists three phases stages. first stage uses 400 historical data rows, second 800 third 1200 rows data. research obtains best result accuracy stage. smallest MSE value: 0.00025248