A Multi-objective Particle Swarm Optimization Framework for Operations Management

There is ongoing research on the problem of how to best combine predictive modeling and optimization. This is especially important in operations management, where there are complex business processes to be optimized. We propose a framework based on evolutionary computing with multi-objective particle swarm optimization and on the design of the fitness function according to the business operations to be optimized. By doing so, one can optimize a range of interesting problems using neural networks that would be otherwise hard to handle with classically supervised learning.