Dynamic Decision Making
Our research focuses on an advanced theory of dynamic decision-making (DM) under uncertainty and incomplete knowledge. This theory aims to strengthen DM processes for both human and artificial agents.
The significance of our theory extends far beyond academic circles, playing a vital role in various contemporary applications, particularly in artificial intelligence. As AI agents navigate uncertain environments, a robust theoretical framework is essential for ensuring effective, reliable outcomes. This theory improves their intelligence and enhances their ability to address complex situations while aligning with specific goals.
Partial solutions derived from this theory have been tested in fields such as medical diagnostics, manufacturing process control, complex robotic systems, economic decision-making, e-democracy, and transportation. These applications highlight the importance of a well-defined decision-making theory in today’s rapidly evolving technological landscape.
Our research is founded on three key pillars:
Experience from real decision tasks
A universality-seeking prescriptive decision theory has to respect that decisions are made in the real world. We continually enhance our expertise in the mentioned fields [1,2].
It reveals: the influence of emotions in economic games [3]; other (ir)rationality aspects [4]; the inevitability of labour division leading to cooperation and negotiation [5], etc., up to the need to use quantum-like modelling of macro-agents [6].
Axiomatic prescriptive decision-making theory
The theory fructifies our experience from real decision tasks and generalises existing theories that focus on the optimisation of expected utility. It models the agent’s environment and decision goals through probabilities [7] and operates solely on them [8].
The original proposal was of a heuristic nature. Gradually, its axiomatic basis was created [9]. It is permanently refined [10]. This research opens new, hard and interesting questions. For instance, a need for a quantum-like variant [11] of the theory.
Mapping of real decision tasks to theory
Users of our theory are experts in their fields and have to focus on their tasks. Thus, they should not be over-loaded by tools that are supposed to help them. Such a need motivates us to systematically build tools that support automatic translation of their task on our theoretical quantitative setup [10].
Importantly, they are constructed within the same theoretical framework. For instance, it helps to tune meta-parameters of algorithms [11] or solves preference elicitation [11]. Classical problems, as decision making with sequential stopping (known as the secretary choice), are also resolved and generalised [13].
[1] Quinn A., Ettler P., Jirsa L., Nagy I., Nedoma P.: Probabilistic advisory systems for data-intensive applications, Int. J. ACASP, 17(2): 133-148, 2003
[2] Kuklišová Pavelková L., Belda K.: Output-Feedback Model Predictive Control Using Set of State Estimates, Informatics in Control, Automation and Robotics. ICINCO 2021 : Revised Selected Papers, p. 151-162, Eds: Gusikhin O., et al, 2023
[3] Guy T.V., Kárný M., Villa A. P., Lintas A.: Theoretical models of decision-making in the Ultimatum Game: Fairness vs. Reason , Advances in Cognitive Neurodynamics (V), 185-191, Eds: Wang Rubin, Pan Xiaochuan, ICCN 2015, Sanya, CN, 2016
[4] Guy T.V., Kárný M., Wolpert D. H. : Decision Making: Uncertainty, Imperfection, Deliberation and Scalability, Springer, Studies in Computational Intelligence vol. 538, 2014
[5] Homolová J., Zugarová E., Kárný M., Guy T.V.: On Decentralized Implicit Negotiation in Modified Ultimatum Game, Multi-Agent Systems and Agreement Technologies, p. 357-369, Eds: Belardinelli F., European Conference on Multi-Agent Systems (EUMAS) 2018
[6] Gaj A., Kárný M: Quantum Model of Uncertainty for Dynamic Decision Making. Quantum Information and Probability: from Foundations to Engineering, Vaxjo QIP24, 2024
[7] Šindelář, J., Vajda, I., Kárný, M.: Stochastic control optimal in the Kullback sense, Kybernetika 44(1):53-60, 2008
[8] Kárný, M., Prescriptive Inductive Operations on Probabilities Serving to Decision-Making Agents. IEEE Transactions on Systems Man Cybernetics-Systems, 52(4):2110-2120, 2022
[9] Kárný M., Kroupa T.: Axiomatisation of fully probabilistic design, Information Sciences 186(1):105-113, 2012
[10] Kárný M.: Axiomatisation of Fully Probabilistic Design Revisited, Systems and Control Letters 141, 2020
[11] Gaj, A.: Quantum Model of Uncertainty for Dynamic Decision Making, 2024
[10] Guy T. V., Kárný M., Wolpert D. H.: Decision Making with Imperfect Decision Makers, Springer-Verlag, Berlin, 2012
[11] Kárný Miroslav: Towards on-line tuning of adaptive-agent’s multivariate meta-parameter, International Journal of Machine Learning and Cybernetics 12(9):2717-2731, 2021
[12] Kárný, M., Guy, T. V., Preference Elicitation within Framework of Fully Probabilistic Design of Decision Strategies, IFAC-Papers on Line: Proceedings of the 13th IFAC Workshop on Adaptive and Learning Control Systems 52(29):239-244, 2019
[13] Karlík D.: Dynamic Fully Probabilistic Decision Making with Stopping, 2024