Institute of Information Theory and Automation

PhD. Topic: Theory and algorithms for probabilistic support of operators (Kárný)

Type of Work:

dissertation

Affiliation/Phone:

ÚTIA AV ČR, v.v.i., oddělení AS, 266052274

Supervisor:

Kárný

Keywords:

Adaptivní systémy, poradní systémy, bayesovské učení, pravděpodobnostní návrh

Hierarchical control and decision making connected with complex processes always contains a layer in which, decsions are made by human being, by an "operator". The proposed topic is related to a group of projects, which aims to create an advanced computer support of such decsion making. Existing original probabilistic theory already proved to efficient for this task. There is however a range of theoretical, algorithmic and software problems that remain to be solved in order to get widely applicable tool. This provides an interesting and useful area for research of 2-3 PhD students.

Bibliography:

[1] Kárný M., Böhm J., Guy T. V., Nedoma P.: Mixture-based adaptive probabilistic control. International Journal of Adaptive Control and Signal Processing, 17 (2003), 2, 119-132. [2]Kárný Miroslav, Böhm Josef, Guy Tatiana Valentine, Jirsa Ladislav, Nagy Ivan, Nedoma Petr, Tesař Ludvík : Optimized Bayesian Dynamic Advising: Theory and Algorithms, Springer, (London 2006) [3]Kárný Miroslav, Guy Tatiana Valentine: Fully probabilistic control design , Systems and Control Letters vol.55, 4 (2006), p. 259-265 [4] Quinn A., Kárný Miroslav, Guy Tatiana Valentine : Fully probabilistic design of hierarchical Bayesian models , Information Sciences vol.369, 1 (2016), p. 532-547

Note:

The topic suits to FJFI, FEL, FIT ČVUT, FAV ZČU and others.

Read more about PhD. Topic: Theory and algorithms for probabilistic support of operators (Kárný)

2022-09-15 10:19

Mgr. Topic: Is the optimal decision making with learning able to win over multi-armed bandits? (Kárný)

Type of Work:

diploma

Affiliation/Phone:

ÚTIA AV ČR, v.v.i., department AS, 266052274

Supervisor:

Kárný

Keywords:

Decision making under uncertainty, Bayesian learning, adaptive control, expoitation and exploration

Probabilistic dynamic systems are appplied in technology, transportation, economics, medicine, electronic democracy etc. They are able to model complex technologies, lumphatic systems or group of automata known as one-arm bandits. Often, structure of the model is known but its parameters are unknown and has to be learnt. Often it has to be done jointly with influencing the system, with control. This creates an interesting and difficult problem as the chosen decisions (inputs, actions) influence both the system and learning efficiency.

Tasks:

1. Learn basics of dynamic decision making under uncertainty. 2. Learn basics of Bayesian learning. 3. Review existing approaches balancing exploitation with exploration. 4. Select or propose the most promising ones and experimentlally verify their properties. 5. Make as general as possible conclusions or hypotheses about inspected decision strategies as the basis for the further research.

Bibliography:

Selected parts: 1.V. Peterka, Bayesian approach to system identification, in P. Eykhoff ed., Trends and Progress in System Identification, p. 239-304, Pergamon Press, Oxford, 1981. 2. M. Kárný, T.V.Guy, Fully probabilistic control design, Systems & Control Letters, 55:4, 259-265, 2006 3. M. Kárný et al, Optimized Bayesian Dynamic Advising: Theory and Algorithms, Springer, London, 2006 4. M. Kárný et al: Dynamic Decision Making" Fully Probabilistic Design, http://www.utia.cz/AS/education/e-materials/main

Note:

Topic at FJFI ČVUT but it can be solved at other faculties or universities.