Forecasting analysis on perak state of consumer price index / Wan Nursyahada Nazurah Meor Muhammad Sulaiman and Faridah Zulkipli

Consumer Price Index (CPI) is important to predict the fluctuation of inflation on each country. The objectives of this study are to study the pattern of component exist in the time series, to determine the most suitable model best fits and to forecast on the step ahead forecast on the
best model. In this study, Perak's state data has been obtained from January 2013 until February 2018. All 62 data were analyzed by using Naïve with trend, Single Exponential Smoothing, Double Exponential Smoothing and Holt’s Method as the error measure are Mean Square Error (MSE) and Mean Absolute Error Percentage (MAPE). As a result, Perak State CPI has a linear trend model and the best model for the time series is Single Exponential Smoothing method with α = 0.99.