I trust you are doing well. My research interests is on forecasting of Tourist Arrivals. I attempted to perform long-term forecasting i.e., from 2019-2031. The data is available from 2008 to 2021, however, the pandemic years 2019-2021 data is not credible for forecasting. I have eliminated 2019-2021 to demonstrate the pre-pandemic normal tourism arrivals in 2031. Used Decomposition, Box Jenkins, and Holt Winter's and validated the accuracy using RMSE, MAPE, and Theil'd U coefficients. However, I have recently received a comment from the research community that "MAPE should not represent the forecasting performances from 2019-2031." Could you please help me understand the reason for this comment as to why MAPE is not suitable for forecast accuracy estimations?
thank you for the video! when i run the CUSUM sq test without log i get a stable results while when i use log the graph is not stable, what should i do in this case? ( PS: i use log because my data numbers are too high, and my model is vecm do i need this test ?)
Thanx for this explanation. Is it possible to test the structural breaks in panel data by Eviews? If yes , how ? Thanx again
Can this process be utilized in the ARDL model. Can you suggest the video pertaining the detection of structural breaks in ARDL?
I trust you are doing well.
My research interests is on forecasting of Tourist Arrivals. I attempted to perform long-term forecasting i.e., from 2019-2031. The data is available from 2008 to 2021, however, the pandemic years 2019-2021 data is not credible for forecasting. I have eliminated 2019-2021 to demonstrate the pre-pandemic normal tourism arrivals in 2031. Used Decomposition, Box Jenkins, and Holt Winter's and validated the accuracy using RMSE, MAPE, and Theil'd U coefficients. However, I have recently received a comment from the research community that "MAPE should not represent the forecasting performances from 2019-2031."
Could you please help me understand the reason for this comment as to why MAPE is not suitable for forecast accuracy estimations?
thank you for the video! when i run the CUSUM sq test without log i get a stable results while when i use log the graph is not stable, what should i do in this case? ( PS: i use log because my data numbers are too high, and my model is vecm do i need this test ?)
Is your data having negative values
Hi Sir can we do structural break analysis on panel data on eviews
No
@@DhavalSaifaleeAaryash thanks for replying sir.