Scientists often run experiments to distinguish competing theories. This requires patience, rigor, and ingenuity—there is often a large space of possible experiments one could run. But we need not comb this space by hand—if we represent our theories as formal models and explicitly declare the space of experiments, we can automate the search for good experiments, looking for those with high expe...

In this contribution we extend a recently developed framework for open loop input design to closed loop experiment design. More specifically, for the very common situation of a fixed controller during the identification experiment, the framework is extended to the design of an optimal reference signal spectrum and also to cope with closed loop signal constraints. In a further extension also the...

This paper develops the idea of min-max robust experiment design for dynamic system identification. The idea of min-max experiment design has been explored in the statistics literature. However, the technique is virtually unknown by the engineering community and, accordingly, there has been little prior work on examining its properties when applied to dynamic system identification. This paper i...

I consider the question "How should one act when the only goal is to learn as much as possible?". Building on the theoretical results of Fedorov (1972, Theory of Optimal Experiments, Academic Press) and MacKay (1992, Neural Computation, 4, 590-604), I apply techniques from optimal experiment design (OED) to guide the query/action selection of a neural network learner. I demonstrate that these t...

Optimal experimental design practices are prominent in many applications. This paper proposes an alternate way of computing the information matrix, a key consideration in planning an accelerated life test. The generalized linear model approach allows optimal designs to be computed using iteratively weighted least-square solutions versus a maximum likelihood method. This approach is demonstrated...

A survey was carried out to characterize experimentation in three different European regions: the Baden-Wurttemberg region, The Basque Country and the rest of Spain. Results of the survey show that even though experimentation is a frequent activity, almost 95% of companies conduct experiments; the strategies used to carry them out are primitive. The one-factor-at-a-time strategy is used by 75% ...