简介: | Often the maximum likelihood estimation of parameters beyond the location-scale setup encounters numerical problems. Cox and Reid (1987) developed a method called the local parameter orthogonality to mitigate the situation. However, they left a few unanswered questions. In this talk, I shall give an affirmative reply to one of them and show how it can lead to efficient algorithms with the celebrated Whittle algorithm for multivariate autoregressive modeling as a showcase. |