Pushing the Limits of the Reactive Affine Shaker Algorithm to Higher Dimensions
This work addresses the challenge of scaling optimization algorithms to very high dimensions for researchers and practitioners in machine learning and optimization, but it is incremental as it builds on existing stochastic search methods.
The paper tackles the problem of optimizing expensive functions in high-dimensional spaces by proposing the Reactive Affine Shaker (RAS) algorithm, which uses a simple stochastic local search method without directly using function values to adapt the search area, and finds that its results are comparable to state-of-the-art high-dimensional Bayesian Optimization, though requiring more function evaluations.
Bayesian Optimization (BO) for the minimization of expensive functions of continuous variables uses all the knowledge acquired from previous samples (${\boldsymbol x}_i$ and $f({\boldsymbol x}_i)$ values) to build a surrogate model based on Gaussian processes. The surrogate is then exploited to define the next point to sample, through a careful balance of exploration and exploitation. Initially intended for low-dimensional spaces, BO has recently been modified and used also for very large-dimensional spaces (up to about one thousand dimensions). In this paper we consider a much simpler algorithm, called "Reactive Affine Shaker" (RAS). The next sample is always generated with a uniform probability distribution inside a parallelepiped (the "box"). At each iteration, the form of the box is adapted during the search through an affine transformation, based only on the point $\boldsymbol x$ position and on the success or failure in improving the function. The function values are therefore not used directly to modify the search area and to generate the next sample. The entire dimensionality is kept (no active subspaces). Despite its extreme simplicity and its use of only stochastic local search, surprisingly the produced results are comparable to and not too far from the state-of-the-art results of high-dimensional versions of BO, although with some more function evaluations. An ablation study and an analysis of probability distribution of directions (improving steps and prevailing box orientation) in very large-dimensional spaces are conducted to understand more about the behavior of RAS and to assess the relative importance of the algorithmic building blocks for the final results.