- #1
mertcan
- 340
- 6
Hi, I have some crucial questions belong to statistics:
First, How can we derive the variance function with respect to mean for a given data?
Secondly, I would like to ask: what method should we employ if the variance in time series behaves like a high order (such as ##𝑎𝑢_𝑡^5+𝑏𝑢_𝑡^4+𝑐𝑢_𝑡^3## polynomial function with respect to mean? On internet, I have always encountered the case where variance is a function of ##𝑢_𝑡^2 or 𝑢_𝑡^4## like in this link I have not seen a case that variance is a function such as ##𝑎𝑢_𝑡^5+𝑏𝑢_𝑡^4+𝑐𝑢_𝑡^3## . What should we do for the last case? How do we find the optimal power transformation or optimal other transformation methods to reduce heteroscedasticity?
First, How can we derive the variance function with respect to mean for a given data?
Secondly, I would like to ask: what method should we employ if the variance in time series behaves like a high order (such as ##𝑎𝑢_𝑡^5+𝑏𝑢_𝑡^4+𝑐𝑢_𝑡^3## polynomial function with respect to mean? On internet, I have always encountered the case where variance is a function of ##𝑢_𝑡^2 or 𝑢_𝑡^4## like in this link I have not seen a case that variance is a function such as ##𝑎𝑢_𝑡^5+𝑏𝑢_𝑡^4+𝑐𝑢_𝑡^3## . What should we do for the last case? How do we find the optimal power transformation or optimal other transformation methods to reduce heteroscedasticity?