Scott R Vandeleest

Title: Scott R. Vandeleest: Inventor in Provider Activity Anomaly Detection.

Introduction

Scott R. Vandeleest is an inventor based in Green Bay, Wisconsin, and is associated with Optum, Inc. His patent work addresses the use of unsupervised learning to identify provider addresses that may have become inactive.

Latest Patents

Vandeleest is named on 1 patent, titled “Provider Activity Anomaly Detection.” The invention describes techniques for detecting inactive provider addresses by analyzing activity trends. When an address shows increasing activity, the system calculates an active metric indicating the likelihood that it remains active. When activity decreases, it calculates an inactive metric indicating the likelihood that the address is inactive.

The system then uses these metrics to classify the address as active or inactive. In some embodiments, the metrics are weighted sums of z-scores derived from Gaussian distributions and benchmarks or provider features. The patent also describes using reinforcement learning from human feedback (RLHF) techniques to determine the weights.

Career Highlights

Vandeleest’s listed patent reflects work at the intersection of data analysis and healthcare provider information management. The approach aims to help identify potentially inactive addresses through automated analysis of activity patterns.

Collaborations

Vandeleest is associated with Optum, Inc. His listed coworkers include Rakesh P. A. and Vijaychandar Natesan.

Conclusion

Scott R. Vandeleest’s patent focuses on applying unsupervised learning and activity metrics to detect potentially inactive provider addresses.