Chicago to Host Fall Seminar on Predictive Modeling

by Ronald T. Kozlowski, Chairperson, Committee on Special Interest Seminars

Learn how insurance entities are using predictive modeling techniques for marketing, underwriting, pricing, and fraud detection at the Seminar on Predictive Modeling, October 4-5 at the Omni Chicago Hotel.

Today, entities of all types are using predictive modeling techniques. Predictive modeling techniques have helped companies identify future customers through analyzing magazine subscriptions, have helped Starbucks determine its next store location based on analyzing demographics, and even helped the government identify people with possible ties to terrorist organizations. The novel use of credit scoring is one of the most successful innovations in personal lines in recent decades. While the appropriateness of using credit scoring for underwriting and rating is still debated, few people can deny the predictive ability of such variables.

The sessions will be designed for understanding practical applications, as well as providing basic and more advanced instruction on predictive modeling techniques and considerations. The seminar will include practical application sessions on predictive modeling for personal auto, homeowners, commercial lines, and medical malpractice; other potential rating variables for personal lines; and predictive modeling for the company with little or no data.

By predictive modeling we are referring to a wide range of techniques from simple multiple regressions to generalized linear models (GLMs), clustering, classification and regression trees (CART), multivariate adaptive regression splines (MARS), and neural networks.

This seminar is perfect for insurance professionals and actuaries of all levels seeking to implement predictive modeling in marketing, underwriting, pricing and fraud detection. Sessions will be geared towards all levels, from actuarial students and Fellows learning predictive modeling basics and advanced techniques to senior management eager to understand predictive modeling capabilities.