# Statistical Inference in Autoregressive Models Год: 2013
Автор: B. Ramanjineyulu,Balasiddamuni Pagadala and G.Mokesh Rayalu
Описание:
In this book,an attempt has been made by developing some inferential methods for autoregressive models by using Internally studentized residuals.In the Applied regression analysis, the autoregressive models, moving average models and combined autoregressive and moving average models have a wide number applications. The study on autoregressive process/models is considered to be essential to both the theoretical and applied statisticians.The first order and higher order autoregressive models for regressed variable and errors have been described by giving auto covariance functions.Further,an autoregressive dynamic model without constant term has been specified and in the presence of lagged dependent variable, a modified durbin’s h-statistic for testing the hypthesis of no auto correlation has been developed for first order autoregressive error process, Instrumental variable method of estimation has been proposed to estimate the parameters of first order autoregressive errors model with…

## 4 Replies to “Statistical Inference in Autoregressive Models”

1. Thank you so much for making short idiot-proof videos with minimal mathematical formulae. Really appreciated (from a doctor who doesnt have the time or patience to sit down and comprehend complicated stats!)

2. Great intuitive explanation!!

3. How do you determine the optimal block size? If you pick bs=1 then it reduces down to the normal bootstrap. So picking the block size must have a big impact on the final result.

1. Good question. For the most part this technique is used with autoregressive models to estimate ranges of parameters etc. One of the hyperparams of those models is how far back you are going to look. Generally speaking matching that hyperparam to the one above is typically done.

I guess you could also do calibration on the block size as well (check out the bootstrap calibration video). But Ive never personally done that myself 🙂