ARIMA(p,d,q) Models (Video 6 of 7 in the gretl Instructional Video Series)
Vložit
- čas přidán 16. 02. 2015
- The gretl Instructional Video Series consists of seven videos that instruct and demonstrate how to use gretl to apply econometric techniques. The videos are designed to be 'hands on' and will be the most effective if the students follow and actively participate using gretl on their own computer while watching the video. The videos can be used for individual study or for in-class presentation.
The learning objectives for Video 6 include the following: ADF unit root test, ACF, PACF, ARIMA model estimation, forecasts, and forecast performance measures.
The data files used in this video can be downloaded from www.janellemann.com/uploads/8/...
Citation: Mann. J. and V. Bindal. (2011, September 1). ARIMA Models - Video 6 of 7 in the gretl Instructional Video Series. Retrieved from • ARIMA(p,d,q) Models (V...
It is important note that the videos are not designed to teach econometric theory and techniques, rather they apply the econometric theory and techniques learned in class.
This is my first time learning about time series forecasting yet this ARIMA lesson has given me so much understanding. Thanks Janelle
they didn't rellay get into how to select p and q, "yeah according to the correlogram it's obviously 1,1" well it's not obvious for me
Hi Janelle, First of all thank you for these great videos. If possible make a series of videos on time series i.e. VAR/VECM and Panel data analysis. Thanks .
I've just loved your videos, they're helping me a lot. Good Job!
This is key, CRUCIAL TO MY LEARNING EXPERIENCE!
Muchas gracias por su video :)
Great video! although I have a question, what if I have independent variables and just want to understand the relationship with the dependent one. By using ARIMA, how should I define the lag and get the (p, d, q)? I am trying to interpret the relationship between Venezuela's GDP Growth and independent variables namely Crude production, Spare Capacity or Economic Freedom from 2001 to 2015 (YoY). This is to answer my RQ of 'What factors hinder economic growth in oil-producing countries?' - Thanks
Hi Janelle,
Thanks very much for the helpful video. I have data that do show non-stationarity with a time trend. I cannot figure out how to determine what value of d to choose for these data. I've seen discussions online suggesting that you need to 'difference' the data series and then test it again, but I also cannot figure out how to do this in Gretl. Could you help point me in the right direction?
you are brilliant
Hello, can you explain why is 1:1 the p and q please?
Thank you, the video is awsome. Does anyone know how can I execute SARIMA (p,d,q)(P,D,Q)s particulary the the last coeficient "s" ?
Hi. How can we account for negative confidence interval in the case of price forecasting using ARIMA?
Hello Anupama Shroff - I suggest double checking that the price series is stationary. Typically price series are first difference stationary, thus you should work with the differenced price series when using the ARIMA model, which would solve your problem.
Thanks for a wonderful presentation. i found this vedio useful. i downloaded the gretl software from google and am forecasting residential estate price using arima. At the end the forecasting model shows some figures with negative sign instead of residential price as it is in my data. so i couldnt understand why and i cannot interpret it. am expecting to see the price in Y axis and years in X axis. Moreover at the end of the forecast it also cannot show forecast evaluation.
Sorry it shows something like this in the graph in Y axis like 2.8e+007 , 2.6e+007, 2.4e+007, 2.2e+007, 1.8e+007..etc. It didnt show me the price of real estate as it is in my data and as i see in your forecasted model. Thanks
Please I need the data you used
While this video is just amazing, for learning purposes, I think it would have been more useful if you picked a series that does need to be differenced to showcase how to do that. :)
Thank you for the HookVonStun. If you know the series is stationary, then you can skip step 1 and start with step 2.
So incomplete :(
CONTACT MEGREAT VIDEO COULD ANYONE HELP ME TO ANALYSIS MY DATA
Hi dear Mann! Thank you for the nice presentation and valuable lectures on grtle. I watched the ARIMA modelling videos along with others, you hae very nicely presented all the material with the best teaching method. Could you please guide me how can I forecast the annual data for future periods. I have data on health care expenditure and GDP per capita and want to forecast these two variables for the period up to 2030 , but i dont know how I may get the future values i.e for 2014, 2015 2016 .......and on. data which i have is from 1981 to 2013. Appreciated if you could support your answers with grtle work file in screen shot.
And if you don't mind and using Skype and can share with me your Skype ID to get help when you have time, or can communicate your email.
Hope for positive reply and kind support at your end.
Also requested to give presentations on panel data modelling in grtle too.
Hanib
Hello Hanib - Glad you found the videos useful. If you want to use ARIMA to forecast the individual series you can follow this video, although your forecasts would likely improve with a cointegration type technique given you want to forecast several years into the future. There are several papers looking at this topic: try searching 'health, gdp, and cointegration' in Google Scholar.
Hi Dear Mann!
Thank you for your reply and sharing a very nice video on forecasting. Really appreciated.
I personally request you please, don't forget me to share any new video or post.
Thank you