The field of optimization concerns iterative procedures for finding the x that extremizes a function f(x). Often an optimization problem comes with several information sources concerning the relation between x and f(x) that one can sample at each iteration, each source incurring a different sampling cost. Multi-Information Source Optimization (MISO) is the problem of how to optimally choose among such a set of information sources during an overall optimization algorithm. The key issue is how to combine information from all those sources while trading off the value of each source's samples and the cost of generating them.
One example of MISO is where the information sources are different computational simulators of the climate’s dynamics, with varying accuracy and varying cost (in terms of how fast they run). In this example the MISO goal is to optimally exploit those simulators to find the climate parameters that give the best fit to observational data, subject to a penalty on total cost incurred. Other kinds of MISO applications involve finding optimal designs of engineered systems. Examples include how best to use a set of simulators of an exascale computer to find the optimal architecture of such a computer, and how best to a set of condensed matter simulators together with laboratory experiments to find the material that optimizes some desired physical properties of the material.
MISO is related to existing work on multi-fidelity optimization, multi-disciplinary optimization, active learning, semi-supervised machine learning, adaptive experimental design, and several other bodies of work. However it extends substantially beyond any of them.
This working group will be a gathering of researchers to discuss approaches to multi-sampler optimization and plan potential collaborations for joint work on it.