Hybrid genetic feature selection and support vector machine for prediction LQ45 index in Indonesia stock exchange

Stock market predictions play a very important role and have attracted a lot of attention, this is because stock
price predictions can bring huge profits in the future by making the right decisions. LQ45 index is one of the most popular
and influential stock indices on the Indonesia Stock Exchange. LQ45 index is an index that measures the price performance
of 45 stocks that have high liquidity and large market capitalization and are supported by good company fundamentals and
adjusted every six months at the beginning of February and August. Stocks with declining performance will be excluded
from the index. Prediction of stock composition in the LQ45 index is an important issue in investment, always attracts the
attention of public investors and academics for research. Prediction LQ45 index will be very useful for investors to be able
to see how the prospects for investing in a company's stock in the future. In order to build a better model to predict the
composition of the LQ45 index effectively and efficiently, we developed a prediction model with a hybrid approach using
genetic algorithms and supporting vector machines to predict which companies will enter and leave the LQ45 index. This
proposed algorithm namely GA-SVM. The results show the proposed algorithm yield excellent performance compared
with PSO-SVM, FS-SVM and BE-SVM and promising results with the accuracy is 93.49%.