Logistic Regression: Developing a Model for Risk Analysis
Logistic regression is a useful statistical technique for developing a prediction model for any event that is binary in nature. A binary event can either occur or not occur. It has only two states which may be represented by 1(occurrence) and 0(nonoccurrence). Logistic regression can also be applied in a situation where the dependent variable has more than two classifications. The logistic regression can either be binary or multinomial depending upon whether the dependent variable is classified into two groups or more than two groups, respectively. In this chapter, the discussion shall be made only for binary logistic regression.