Stochastic optimization models for workforce planning, operations, and risk management

Abstract

A framework for unified decision making under uncertainty that supports financial planning, operations management, and risk management for workforce applications is proposed and analyzed. The management of enterprise workforce is conducted at the granularity of cohorts of individuals with similar attributes of interest. A time inhomogeneous Markov chain is developed to model the evolution of these cohorts over time. Stochastic control problems based on versions of the controlled Markov chain are formulated to maximize profit under a set of workforce decisions. Extensive data analysis and innovative computational approaches enable us to solve these stochastic control problems for large-scale systems, with real-world business case studies demonstrating the use of this unified decision support capability for large-scale enterprises.