Penerapan Metode Adaptive Neuro Fuzzy Inference System (ANFIS) Dengan Membership functionTipe Gaussian dan Generalized BellDalam Prediksi Harga Tertinggi Saham

Many people who have capital are currently buying up shares in the stock market in the hope that the stock price will rise when the Covid-19 pandemic ends. As someone who wants to try investing in the stock market, you must be able to estimate the profits and losses from buying shares. One way that can help consideration in mak ing decisions to buy and sell shares is to mak e predictions. There are many algorithms that can be used in prediction, one of which is the Adaptive Neuro Fuzzy Inference System (ANFIS) method which is a combination of the Fuzzy Logic algorithm and Artificial Neural Network s. The application of the ANFIS method requires a good ANFIS structure by selecting the right number and type of membership functions. In this study the Gaussian and Gbell type membershipfunctions are used because they have the advantage of allowing subtle changes and can accommodate inaccuracies in measurements so that they match the pattern of historical data that moves smoothly at one time. In this study, it was found that the gaussiantype has better accuracy than the gbell type by 97.87% to predict the highest stock price of Tencent Holdings Limited and the gbell type has a better accuracy than the gaussian type of 97.8% to predict the highest price of Take-Two shares. interactive.Keywords: Prediction, ANFIS, Membership function, Stock, Data.