Our World Statistics Day conversations have been a great reminder of how much statistics can inform our lives. Do you have an example of how statistics has made a difference in your life? Share your story with the Community!
Choose Language Hide Translation Bar
Highlighted
vmaynard10
Level I

Error with AR model: Cannot Decrease Objective Function Hessian is not positive definite

I am getting the following error while producing a ARI(1,1) model. The steps I have followoed to produce the model below are: 

1. Analyze > Specialized Modeling > Time Series
2. Select appropriate x and y.
3. Change Autocorrelation Lags to 55
4. Once plotted, click red down arrow near the title and select "Difference". Set Nonseasonal Plot to 1.
5. Once plotted, click red down arrow near the title and select "ARIMA". Change p.Autoregressive Order to 1 and Differencing Order to 1. Click Estimate.

 

I am sure I am following the correct steps. Can anyone help me resolve this issue? 

jmp.png

1 REPLY 1
Highlighted
ih
ih
Level VII

Re: Error with AR model: Cannot Decrease Objective Function Hessian is not positive definite

Is it possible that the data cannot be reliably modeled using a time series?  One way to check this is to use a validation set, for example exclude all of the values after 2016, build a model, and try to predict 2017.

 

A web search for 'positive hessian Matrix' will yield some very technical and in-the-weeds articles, but this example might help.  Basically, it means the model isn't able to build a model that it think you should trust.

 

https://www.theanalysisfactor.com/wacky-hessian-matrix/

Article Labels