Loglinear regression is a kind of regression aimed at finding the best fit between the data and a loglinear model.
The major assumption of loglinear regression is that a linear relationship exists between the log of the dependent variable and the inependent variables.