ARIMA Model Forecasting issue

4 次查看（过去 30 天）
Michael Mueller
Michael Mueller 2019-3-1
评论： Stephen Gray 2025-4-27
I am trying to creat an ARIMA model to predict future data points. Using the Econometrics App I fitted a model ARIMA(4,1,3) that fits the data very well. I then try to forcast the data using forcast.
SteelData1Log = log(SteelData1);
SteelData1LogDiff = [NaN; diff(SteelData1Log)];
ARIMA_SteelData1LogDiff2 = arima('Constant',NaN,'ARLags',1:4,'D',1,'MALags',1:3,'Distribution','Gaussian');
ARMA = estimate(ARIMA_SteelData1LogDiff2,SteelData1LogDiff,'Display','off');
yF1 = forecast(ARMA,44,'Y0',August2017.Var2(1:100));
The result does not follow the data, and has a logorithmic trend to it.
  2 个评论
BN
BN 2022-11-18
Dear @Michael MuellerMichael,
Did you find an answer for this problem?
Stephen Gray
Stephen Gray 2025-4-27
I too am looking for an answer to this!

请先登录，再进行评论。

请先登录，再回答此问题。

回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Conditional Mean Models 的更多信息

标签

产品


版本

R2018b

Community Treasure Hunt

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

Start Hunting!

Translated by