Hi! I'm doing a time series analysis over a stationary series. Hence I used the ARMA Model. But I'm getting positive SIC values. Even though I got good p-values (signifying a good fit) but I just wanted to be sure if its normal to get +ve SICs.
It would be great if anyone could give their valuable inputs at the earliest. Thank you!
do you mean Standard Industrial Classification (SIC) Values?
please give us more explanation about this point.
Schwarz Information Criterion (SIC)
See these links:
https://www.stat.ipb.ac.id/en/uploads/KS/S2%20-%20ADW/3%20Montgomery%20-%20Introduction%20to%20Time%20Series%20Analysis%20and%20Forecasting.pdf
..... page 59.
https://en.wikipedia.org/wiki/Bayesian_information_criterion
Thank you so much for your kind help, Prof Remal Al-Gounmeein and Prof Ette Etuk
Marwa Ghareeb Ghanem Ma'am, by SIC I meant Schwartz Information Criterion.
It may depend on the formula used to calculate it. You can either use the full log-likelihood (including the 2pi factor and the estimate of the variance), the concentrated likelihood, minus the sum of squares or the sum of squares.
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series