Towards Adaptive Enterprise: Adaptation and Learning

The key to a successful adaptive enterprise lies in techniques and algorithms that enable the enterprise to learn about its environment and use the learning to make decisions that maximize its objectives. The volatile nature of the contemporary business environment means that learning needs to be continuous and reliable, and the decision-making rapid and accurate. In this chapter, the authors investigate two promising families of tools that can be used to design such algorithms: adaptive control and reinforcement learning. Both methodologies have evolved over the years into mathematically rigorous and practically reliable solutions. They review the foundations, the state-of-the-art, and the limitations of these methodologies. They discuss possible ways to bring together these techniques in a way that brings out the best of their capabilities.