Keynotes
The Logic of Causality
Abstract
The collection of massive observational datasets has led to unprecedented opportunities for causal inference, such as using electronic health records to identify risk factors for disease. However, our ability to understand these complex data sets has not grown the same pace as our ability to collect them. Instead causal inference has mainly focused on identifying pairwise relationships between variables. However, this leads to a large network of relationships that can be difficult for non-experts to make sense of, and further can lead to ineffective actions by obscuring important details of the relationship. In this talk I discuss how temporal logic can be used to represent and test far richer relationships, leading to the ability to find how the duration of a cause, or combination of variables leads to a particular effect. We also show how temporal logic can be used to distinguish between causes and moderators of a causal relationship. Finally, we examine the problem of explanation for specific events, such as why an individual’s glucose is raised.
Bio
Samantha Kleinberg is an Assistant Professor of Computer Science at Stevens Institute of Technology. She received her PhD in Computer Science from New York University in 2010 and was a Computing Innovation Fellow at Columbia University in the Department of Biomedical informatics from 2010-2012. She is the recipient of NSF CAREER and JSMF Complex Systems Scholar Awards and her work is also supported by the NIH through an R01. She is the author of Causality, Probability, and Time (Cambridge University Press, 2012) and Why: A Guide to Finding and Using Causes (O’Reilly Media, 2015).
Causality and temporal dependencies in the design of fault management systems (extendend abstract)
Abstract
Reasoning about causes and effects naturally arises in the engineering of safety-critical systems. A classical example is Fault Tree Analysis, a deductive technique used for system safety assessment, whereby an undesired state is reduced to the set of its immediate causes. The design of fault management systems also requires reasoning on causality relationships. In particular, a fail-operational system needs to ensure timely detection and identification of faults, i.e. recognize the occurrence of run-time faults through their observable effects on the system. Even more complex scenarios arise when multiple faults are involved and may interact in subtle ways.
In this talk, I will present a formal approach to fault management for complex systems. I will introduce the notions of fault tree and minimal cut sets. I will then present a formal framework for the specification and analysis of diagnosability, and for the design of fault detection and identification (FDI) components. Finally, I will review recent advances in fault propagation analysis, based on the Timed Failure Propagation Graphs (TFPG) model.
Bio
Marco Bozzano is a senior researcher at Fondazione Bruno Kessler, Trento (Italy). He received his PhD degree in Computer Science from the University of Genova in 2002. His research interests include formal methods, model checking, formal safety assessment and fault management systems. He has co-authored more than 60 papers on these topics, and one book titled "Design and Safety Assessment of Critical Systems". In recent years, he has been involved in several technology research projects funded by the European Space Agency on the topics of dependability analysis and fault management.