PREDIKSI HARGA EMAS DENGAN MEMBANDINGKAN ALGORITMA NAÏVE BAYES, K-NEAREST NEIGHBOR DAN SUPPORT VECTOR MACHINE UNTUK MEMINIMALKAN RESIKO INVESTASI

Abstract

There are several types of investments to make money, one of which is with precious metal gold investments. Gold investment movements tend to be more stable and increase in value. In addition, gold is a form of liquid investment, which means it can be accepted in any region or country and easy to do disbursement in a short time. When the potential return on investment in stocks or bonds is no longer attractive and is considered unable to compensate for existing risks, then the investor will divert funds to tangible assets such as precious metals or properties that are considered more feasible and safer. To avoid a lot of investment losses, we need an analysis model of gold price movements, in this study the researchers chose to use the Naïve Bayes Algorithm, K-NN and SVM. Based on the background of the problem described earlier, the aim of this study is to obtain a competitive prediction analysis model for the level of accuracy and the level of RMSE from the data set obtained from www.fianance.yahoo.com in the 5- period of May 1, 2014 - May 1, 2019 and will be tested using different test data ranging from 1 month, 3 months, 6 months, 1 year, 3 years and 5 years. After conducting research using models from the Naïve Bayes, KNN and SVM algorithms, we can conclude that the more data used, the higher the accuracy, the truer predictions, and the lower RMSE.