A generalized linear model (GLM) is a multiple linear regression equation whose results can be generalized to wider population than that from which the sample used to estimate its parameters was selected.
The link function of a logistic regression model is the logit of the rate.
The link function of a Poisson regression model is the natural logarithm of the rate of the event.
When the outcome of interest is a continuous numerical variable, the link function for the GLM commonly used in these circumstances is the identity link.
The method of least squares is generally used to estimate the parameters of a generalized linear model.