April 2024
Intermediate to advanced
264 pages
6h 10m
English

When is it preferable to use Poisson regression over ordinal regression, and vice versa?
We usually use Poisson regression when the target variable represents count data (positive integers). As an example of count data, consider the number of colds contracted on an airplane or the number of guests visiting a restaurant on a given day. Besides the target variable representing counts, the data should also be Poisson distributed, which means that the mean and variance are roughly the same. (For large means, we can use a normal distribution to approximate a Poisson distribution.)
Ordinal data is a subcategory of categorical ...
Read now
Unlock full access