Learning to Set Up Numerical

Gradient-based numerical optimization of complex engineering designs ooers the promise of rapidly producing better designs. However, such methods generally assume that the objective function and constraint functions are continuous , smooth, and deened everywhere. Unfortunately , realistic simulators tend to violate these assumptions, making optimization unreliable. Several decisions that need to be made in setting up an optimization, such as the choice of a starting prototype, and the choice of a formulation of the search space, can make a diierence in the reliability of the optimization. Machine learning can improve gradient-based methods by making these choices based on the results of previous optimizations. This article demonstrates this idea by using machine learning for four parts of the optimization setup problem: selecting a starting prototype from a database of prototypes, synthesizing a new starting prototype, predicting which design goals are achievable, and selecting a formulation of the search space. We use standard tree-induction algorithms (C4.5 and CART). We present results in two realistic engineering domains: racing yachts, and supersonic aircraft. Our experimental results show that using inductive learning to make setup decisions improves both the speed and the reliability of design optimization.