0473911 20240103213953.220170412235959.9 85015609196 000400231600008 10.1016/j.ijar.2017.02.001 9 s. P cav_un_epca*02567740888-613X841 (2017)150-158Elsevier Fully probabilistic design Parameter prior External predictive distribution Bayesian transfer learning Kullback–Leibler divergence cav_un_auth*0021112 Quinn A. IE cav_un_auth*0101124 Adaptivní systémy Department of Adaptive Systems AS AS Department of Adaptive Systems Kárný Miroslav UTIA-B Ústav teorie informace a automatizace AV ČR, v. v. i. cav_un_auth*0101092 Guy Tatiana Valentine Adaptivní systémy Department of Adaptive Systems AS AS UTIA-B Department of Adaptive Systems Ústav teorie informace a automatizace AV ČR, v. v. i. http://library.utia.cas.cz/separaty/2017/AS/guy-0473911.pdf cav_un_auth*0331019 GA16-09848S GA ČR CZ This paper exploits knowledge made available by an external source in the form of a predictive distribution in order to elicit a parameter prior. It uses the terminology of Bayesian transfer learning, one of many domains dealing with reasoning as coherent knowledge processing. An empirical solution of the addressed problem was provided in [19], based on an interpretation of the external predictor as an empirical distribution constructed from fictitious data. In this paper, two main contributions are provided. First, the problem is solved using formal hierarchical Bayesian modeling [25], and the knowledge transfer is achieved optimally, i.e. in the minimum-KLD sense. Second, this hierarchical setting yields a distribution on the set of possible priors, with the choice [19] acting as the base distribution. This allows randomized choices of the prior to be generated, avoiding costly and/or intractable estimation of this prior. It also provides measures of uncertainty in the prior choice, allowing subsequent learning tasks to be assessed for robustness to this prior choice. The instantiation of the method in already published applications in knowledge elicitation, recursive learning and flat cooperation of adaptive controllers is recalled, and prospective application domains are also mentioned. BC 10000 10200 10201 2018 3 4 A hod 4ah 20231122142404.6 RVO:67985556 http://hdl.handle.net/11104/0271347 cav_un_auth*0345684 Trinity College Dublin, the University of Dublin TCD IE 1 Department of Adaptive Systems UTIA-B 10201 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE S 2 Article|Proceedings Paper Computer Science Artificial Intelligence 2 Article|Proceedings Paper Computer Science Artificial Intelligence 2 Article|Proceedings Paper Computer Science Artificial Intelligence COMPUTERSCIENCE.ARTIFICIALINTELLIGENCE 2.504 0.687 7.8 0.0042 0.658 3384 0.866 1.343 182 Q2 44.33 51.1 Q3 Q2 0.9 Q2 51.136 2017 Soubory v repozitáři: guy-0473911.pdf 85015609196 SCOPUS PUBMED 000400231600008 WOS cav_un_epca*0256774 International Journal of Approximate Reasoning 0888-613X 1873-4731 Roč. 84 č. 1 2017 150 158 Elsevier