The Event
Schedule & Speakers
Venue & Hotel
Sponsors

Breakout C: Data Sculpting: A New Approach to Creating Robust Simulations for Stress Testing & Score Performance Validation

12:00 PM - Monday, September 23

Data-driven simulations are critically important for evaluating the expected performance of credit portfolios, models, and credit scores. Existing approaches to obtaining simulation input data are often logistically burdensome, technically complex, or overly simplistic. In comes “Data Sculpting”, a novel technology developed in the FICO® Scores Innovation Lab that uses advanced optimization techniques to simulate consumer and portfolio data that is realistic and tunable to specific user-defined scenarios. We’ll demonstrate how a major lender uses Data Sculpting to evaluate the performance of a new score on their through-the-door population.

Speakers

Christopher Ralph

Linkedin
Christopher Ralph
Principal Scientist, FICO® Scores Research, FICO

Chris Ralph is a Principal Scientist in the Scores Research group at FICO, where he leads the design and development of Python-centric machine learning software solutions. His current area of focus is Data-Centric Artificial Intelligence (AI), leading a project to build an extensible, modular, and scalable “Augmented Intelligence for Feature Discovery” solution to maximize the value of new and alternative data sources for all FICO Score products. He also currently leads the FICO Scores Synthetic Data initiative focused on researching new solutions to enable data portability, improve data security, and enhance model robustness using novel augmentation & synthetic simulation techniques.

Chris’s professional experience spans a wide range of increasingly senior roles in Applied Research and Development, Analytic Consulting, Global Analytics Delivery, Applied Optimization, and Scores Advanced Analytics. He has led teams of data scientists, machine learning engineers, software developers, and analytic consultants in the development of modular software frameworks and related end-client solutions for mortgage, loan and deposit pricing optimization, marketing offer optimization, strategic segmentation, and credit risk modeling. He is a named inventor on multiple FICO patents and holds a BEng, Industrial Engineering and Operations Research, from Toronto Metropolitan University, and MBA in Information Systems, McMaster University.