Prediction of Price of Local Fruits in Jakarta With ANFIS (Adaptive Neuro Fuzzy System Infrence)

Abstract

ANFIS (Adaptive Neuro Fuzzy Infrence System) is one of prediction methods that use principles of artificial neural networks and fuzzy logic has been widely used in various disciplines and produce a good prediction. Therefore, the ANFIS method, it can be applied to predict stock prices of fruits and Jakarta, so it can be used to increase the prices of local fruits. Data used for tested in this study is the monthly rainfall data in DKI Jakarta and the cities of local fruit supplier in Jakarta, and price data of fruits in the year 2000 to 2009. From the data analysis were performed PCA (Principal Component Analysis) to reduce the complexity of large amount of data. Analysis of data using PCA and ANFIS prediction made in the form of two dimensions. The length of the data used is 120 months, with as many as 110 data training data, with a predicted 10 months in the future. Validation is done by comparing the predicted data and the original data in the same months.