Naphtha's price forecasting using neuro-fuzzy system

Abstract

Naphtha's price forecasting using Neuro-fuzzy system is a forecasting technique that applied information technology with statistics. 1950 daily prices were collected as a time-series data with trend component. The research found that Neuro-Fuzzy system is more accurate and more reliable than a statistical method; it also works well with continuous data and performs better with more training data. Neuro-Fuzzy system might be used with different data type, but, it might come across with other factors, e.g. seasonal or irregular event. The research also illustrates the multidisciplinary nature in today's world of works in the era of merging among many disciplines. © 2008 SICE.