1

I need to predict the 5 year interest rate from 2010 to 2019 based on data of 1964 to 2010.

I have done forecasting.

I want to improve my RMSE accuracy.

How can I do it ?

My R code is the following

us<-ts(train$X5.year,start = min(train$time_stamp),end = max(train$time_stamp))
usOPT<-auto.arima(us)
usOPT
coef(usOPT)
fit<-predict(usOPT,n.ahead = 3000,se.fit = T)
schlebe
  • 3,387
  • 5
  • 37
  • 50
user10747759
  • 19
  • 3
  • Hello new user. I have added new line to your post and I have highlighted some technical terms. You can see my change in editing your post. Using `word` technic is interesting to render your question more readable. – schlebe Jan 23 '19 at 06:49
  • Can you help me with the question as well? – user10747759 Jan 23 '19 at 06:52
  • No, I cannot help you to find a solution because I don't know R language. I have more than X reputations and I have gain the possibility to check question of new users. I have only accepted your question and make comments to help you to write correct question. When you will have more reputation, you question will be directly accepted but this is not the case for first questions. I have voted up to your question because I think it is a good question even I cannot resolve it. – schlebe Jan 23 '19 at 07:00
  • You could try different forecasting models and evaluate the forecasting performance based on the RMSE. Have a look [here][1] for an exhaustive list [1]: https://stats.stackexchange.com/questions/312320/alternatives-to-using-arima-for-forecasting – Martin Jan 23 '19 at 07:14
  • To improve your chances of getting an answer, please read [this](https://stackoverflow.com/help/how-to-ask) post. In particular, including a [minimal reproducible example](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) in your question is essential. – Samuel Jan 24 '19 at 08:47

0 Answers0