Objective
Automobiles come to market with a fixed price plus additional government taxes. That price is set by the manufacturer based on features like engine, body-style and fuel type, and brand value also matters. Vehicles are built with a budget in mind: the higher the budget, the better the features, the stronger the customer appeal, and these factors all feed into price.
Every car and light commercial vehicle, new or old, falls into a particular insurance group which dictates how much it costs to insure. Insurers use these groups to quantify the risks associated with vehicles, with higher cost contributions tied to higher group numbers. Cars are initially assigned risk-factor groups (“symbols”) tied to price, then adjusted for risk, the higher the risk, the higher the factor. Actuaries call this “symboling”.
That raises the question of what factor influences a car’s risk, whether those factors affect price, and whether price can be predicted by planning ahead on the features included. Answering these questions matters from a company perspective. With that motivation, the objective is to analyse the UCI Automobile dataset (collected from the 1985 Ward’s Automotive Yearbook).