A generalized linear model (GLM) is a flexible extension of ordinary linear regression. A widely used GLM is binary logistic regression. Generally speaking, a GLM consists of a random component and a systematic component:
The following table summarized the available distributions (also called families) and link functions, as well as the suitable type of response data. The asterisk * indicates the canonical/default link function for a specific family.
| Family | Links | Response Data | |------------------|-----------------------------------------------------|---------------------| | Bernoulli | Logit, Probit, Cauchit, Complementary Log-Log, Log | Binary | | Binomial | Logit, Probit, Cauchit, Complementary Log-Log, Log | Proportions, Counts | | Gaussian | Identity, Log, Inverse | Continuous | | Gamma | Identity, Log, Inverse | Positive continuous | | Inverse Gaussian | Identity, Log, Inverse, 1/mu^2 | Positive continuous | | Poisson | Identity, Log, Square-root | Counts |
Dunn, P. K., & Smyth, G. K. (2018). Generalized linear models with examples in R. New York: Springer.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.