Another form of regression analysis is Poisson regression. This type of analysis is a generalized linear model or GLM, used to model count data.

Unlike the example in the previous section where wine samples had been rated (or ranked) on a scale from 1 to 10, count data is a (statistical) data type in which the observations can take only the non-negative integer values {0, 1, 2, 3, ...}, and where these integers arise from counting rather than ranking.

Poisson regression assumes the outcome of your analysis has a Poisson distribution – in that it expresses: the probability of a number of events occurring in a fixed interval of time if these events occur with a known average rate and independently of the time since the last event.

An example model might be of the number of phone calls received by a software support center each hour. Predictors of the number of calls received include the number of days after a new version (of the software) has been released and the number of years the customer has been a user of the software (in keeping with our wine example, you could use Poisson regression to analyze the number of bottles of wine sold during a month, with perhaps store location and the month of the year as predictors).

A data scientist may also use a Poisson distribution for the number of events in other specified intervals, such as distance, area, or volume.