MS17 Decision Making under Uncertainty

Prof. Daniel Straub: straub@tum.de

Session Chairs:
Daniel Straub, Prof., Technical University of Munich, straub@tum.de
Kostas Papakonstantinou, Prof., Penn State University, kpapakon@psu.edu
Matteo Pozzi, Prof., Carnegie Mellon University, mpozzi@cmu.edu
Iason Papaioannou, Dr., Technical University of Munich, iason.papaioannou@tum.de

Abstract of the special session:
Results from every engineering analysis ultimately serve as decision support. Hence, it is often relevant to set the analysis in the context of a formal decision framework, to ensure proper interpretation of results and optimal choices. In addition, the advent of increasingly automated and autonomous systems with independent decision making capabilities requires new approaches and methods for quantifying and validating safety and reliability.
This mini-symposium aims at gathering researchers interested in developing and applying methods related to optimal decision making under uncertainty for engineering systems. It focuses on the integration of formal decision analysis with stochastic models, data analytics, and artificial intelligence methodologies towards optimal decision making in complex settings. These can include, among others, sequential decision processes, and decisions involving multiple objectives and/or stakeholders. Selection of metrics for stochastic analysis based on decision-theoretic considerations are also of interest, such as the choice of appropriate objective functions and decision-theoretic sensitivity measures. Areas of interest, both in terms of methodologies and applications, include, but are not limited to:

• Machine learning / artificial intelligence methods
• Markov Decision Processes under perfect and partial information
• Value of Information and structural health monitoring
• Stochastic control
• Deep learning architectures
• Sensitivity analysis
• Multi-agent systems
• Multi-objective optimization
• Autonomous systems
• Life-cycle infrastructure management and optimization
• Inspection and maintenance planning
• Early warning, resilient recovery and emergency response
• Risk and reliability assessment
• Probabilistic performance-based engineering