W = f( p, q, R ) where, w and p means monthly wages and food prices which are available from 2000 to 2015 from 20 districts (places).
But others like production quantity of rice (q) are seasonal and remittance (R) are in monthly at national level.
How can I develop a time series model on such varied frequency variables.
You can use a VAR model with varied frequency data in a Bayesian context.
You can also use
MIxed Data Sampling (MIDAS) models, mixed frequency VARs, and mixed
frequency factor models
You can easily use various autoregressive models. I suggest you take a look on econometric books.
Yours,
Ehsan
To my view, an efficient way to formaly model such a type of problem is to consider Bayesian hierarchical modeling. I suggest you to read "Applied Bayesina Hierarchical Methods" by Gongdon or Banerjee, S., Carlin, B. P., & Gelfand, A. E. (2004). Hierarchical modeling and analysis for spatial data. 2e éd., Chapman & Hall, Boca Raton, USA, 562 pages.
Best regards,
Luc
I would concur with Luc. Spatial models have been developed by Noel Cressie and Christopher Wikle, amongst others, regarding which may be fit with Hierarchical Mixed Models (see Anthony Bryk and Steve Radebusch, Sage Publication, 2002).
There is a vast literature on these things as Bayesian Disease Mapping: Hierarchical Spatial analysis by Andrew Lawson (2006) and others. The USGS makes use of this material in much of their analysis as does the CDC.
Regards,
Robert Yaffee
Dear All Thanks a lot for your valuable comment.
Dear L.Perreault,
Unfortunately, I am not able to get the hard copy as well as the soft copy of the book you suggested (Applied Bayesian Hierarchical Methods" by Gongdon or Banerjee,). It is available on amazon but it's costly. Can you help me in this regard?
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis