Multivariate classification

Question 1

Benefits of multivariate methods

Answer 1

1. Consider all rating variables simultaneously and automatically adjust for exposure correlations between rating variables
2. Remove unsystematic effects in the data (i.e. noise) and capture only systematic effects (i.e. signal)
3. Model diagnostics
4. Allow consideration of the interaction between rating variables

Question 2

Cluster analysis

Answer 2

An exploratory data analysis tool that seeks to combine small groups of similar risks into larger homogeneous categories or “clusters”

Question 3

Factor Analysis

Answer 3

principle components analysis are most commonly used, a technique to reduce the number of parameter estimates in a classification analysis (e.g. GLM)

Question 4

CART

Answer 4

Classification and Regression Trees

Question 5

MARS

Answer 5

Multivariate Adaptive Regression Splines

Question 6

Neural Networks

Answer 6

Gather test data and invokes training algorithms designed to automatically learn the structure of the data. (recursion applied to a GLM)