Optimal search

There is no single answer to what constitutes 'optimal search.'[1] The definition depends on context, search task, time constraints, and goals. But we can narrow the field when considering the focus of our research. Here, it is time-limited, complex search tasks requiring attention and effort in finding an ill-defined target. Perusng my web pages will lead you to my view that optimal search -- when defined in these terms -- may be achieved through a cellular-automata (CA) type mechanism that captures known properties of human visual system. Interestingly, CA's, as well as human search, often produce haphazard behavior. But paradoxically, this complex behavior can effectively cover the search space to detect targeted information. While there are many CA rules that can serve as candidate models, use of an 'evolutionary design' approach (where iterating CAs through neural-net simulations until match to solution is achieved) along side systematic behavioral tests might hone us in on plausible models of effective search.

------------------------------------------------------------------------------------------------------------------

[1] Bayesian (network) approach, offers a useful and, arguably, universal approach to deriving optimal search. I am will be elaborating on its value along with some of its limitations.