Environmental adaptation and differential replication in machine learning

Abstract

When deployed in the wild, machine learning models are usually confronted withan environment that imposes severe constraints. As this environment evolves, so do these constraints.As a result, the feasible set of solutions for the considered need is prone to change in time. We referto this problem as that of environmental adaptation. In this paper, we formalize environmentaladaptation and discuss how it differs from other problems in the literature. We propose solutionsbased on differential replication, a technique where the knowledge acquired by the deployed modelsis reused in specific ways to train more suitable future generations. We discuss different mechanismsto implement differential replications in practice, depending on the considered level of knowledge.Finally, we present seven examples where the problem of environmental adaptation can be solvedthrough differential replication in real-life applications.