The website is in Maintenance mode. We are in the process of adding more features.
Any new bookmarks, comments, or user profiles made during this time will not be saved.

Machine Learning Resources

What about cases where a significant number of observations have a count of 0 (in the context of Poisson Regression)?

Bookmark this question

In many data generation processes for count data, it is possible that a lot of observations will have a count of zero. For example, if the outcome is something like the frequency of natural disasters, it is possible that in many time windows, no such events occur. When this is the case, it is common to use a Zero-Inflated Poisson model (ZIP).

The ZIP model extends the regular Poisson model by introducing a second component that is used to model the probability of a non-zero count occurring. If the count is non-zero, the regular Poisson distribution is used, and if the count is zero, zeroes are generated at a rate corresponding to the probability of observing a zero count in the data. The ZIP model is an example of a mixture model, where the data generation process is comprised of more than one distribution with a probability assigned to the likelihood that an observation comes from each component. 

Leave your Comments and Suggestions below:

Please Login or Sign Up to leave a comment

Partner Ad  

Find out all the ways
that you can

Explore Questions by Topics

Partner Ad

Learn Data Science with Travis - your AI-powered tutor |