I am developing time series model. Working in daily prediction using "ts" from "library (tseries)".
I use "Holt winters" and "tbats" method but the value seems to be not accurate.
any idea for a best fit to a daily prediction. The code am using..
td<-ts(mydata1_subset$sales, frequency=365, start=c(2018, 1),end = c(2020,200)) past<- window(td, end = c(2020,200)) future<- window(td, start = c(2020,200)) model1<- HoltWinters(past,seasonal = "additive") pred<- predict(model1,n.ahead = 165,prediction.interval = T)
I want to predict daily data for quarterly aggregation.