In this keynote talk, we will survey and explain the state-of-the-art concerning the Stochastic Search on the Line (SSL) problem, also synonymously known as the Stochastic Point Location (SPL) Problem. The SPL was introduced by Oommen in [10], and it has been studied and analyzed by numerous researchers during the last two decades. It involves determining an unknown 'point' when all that the learning system stochastically knows is whether the current point that has been chosen is to the left or the right of the unknown point.In this talk, we will explain how the SPL is a fundamental problem in machine learning, optimization and control, and demonstrate that it is also central to the field of AI. The talk will survey the various automata-based and hierarchical techniques that have been used to solve it, including learning from a Stochastic Teacher or a Compulsive Liar, and in symmetric mechanisms. We will then describe how it is all-pervasive in a variety of application domains and discuss these applications.

Additional Metadata
Keywords Artificial intelligence, Hierarchical learning, Learning automata, Stochastic point location
Persistent URL dx.doi.org/10.1109/ICTCS.2017.70
Conference 2017 International Conference on New Trends in Computing Sciences, ICTCS 2017
Citation
Yazidi, A. (Anis), & Oommen, J. (2018). The theory and applications of the stochastic point location problem. In Proceedings - 2017 International Conference on New Trends in Computing Sciences, ICTCS 2017 (pp. 333–341). doi:10.1109/ICTCS.2017.70