Problem with Arima model forecasting in Matlab 2012a

1 view (last 30 days)
I am using Matlab 2012a and want to forecast the values of a time series.My observations are 16000 and their values range from 0 to 0.15 and their periodicity equals 24. For this purpose I use the Matlab commands arima to define the model,estimate to calculate the parameters and forecast to forecast the values.My problem is that regardless the combination of (p,q) of the arima model and regardless the features of the times series(I removed the seasonality and made the time series stationary according to Matlab's tutorial), the forecast results deviate a lot from the given observations. I have also tried multiplicative arima model giving seasonality parameter 24 in order the SAR and SAM parameters to be estimated.In this case the forecast result seems better,but the overall result can't be considered real.For instance, if the forecast points are set to 1000, the forecast results at intervals 1-24,25-48 ... 977-1000 will be identical. What am I doing wrong?

Answers (0)

Categories

Find more on Conditional Mean Models in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!