Recent advancements within chain graphs
Probabilistic graphical models are currently one of the most commonly used
architectures for modelling and reasoning with uncertainty. They are used in a
wide range applications ranging from error diagnostics in printers to modelling
protein structures in bioinformatics and decision support systems in market
The most commonly used subclass of the models, Bayesian networks, do however have a major limitation, which is that only asymmetric relationships, namely cause and effect relationships, can be modelled between the variables. A class of probabilistic graphical models that tries to address this shortcoming is chain graphs, which include two types of edges in the models representing both symmetric and asymmetric relationships between the variables.
In his thesis, Dag Sonntag discusses the semantics of chain graphs and present recent advancements made in the research field. He also compares the expressivity of the subclass as well as proposes new structure learning algorithms to learn chain graph models from data.
Are your mobile apps consuming too much energy when accessing the network?
Energy consumption and its management have been clearly identified as a
challenge in computing and communication system design, where energy economy is
obviously of paramount importance for battery powered devices.
Ekhiotz Jon Vergara's PhD thesis addresses the energy efficiency of mobile communication at the user end in the context of cellular networks.
The first contribution is EnergyBox, a parametrised tool that enables accurate and repeatable energy quantification at the user end using real data traffic traces as input. EnergyBox models the energy consumption characteristics of wireless interfaces and allows to estimate the energy consumption for different operator settings and device characteristics. EnergyBox facilitates energy consumption studies, and the thesis proposes different energy-efficient solutions from application and operating system perspective. Spotify has used EnergyBox to reduce the energy consumption of their Android application.
The thesis also studies the generic problem of determining the contribution of an entity (e.g., application) to the total energy consumption of a given system (e.g., mobile device). It compares the state-of-the-art policies in terms of fairness leveraging cooperative game theory and analyse their required information and computational complexity. The thesis shows that providing incentives to reduce the total energy consumption of the system (as part of fairness) is tightly coupled to the policy selection and provides guidelines to select an appropriate policy depending on the characteristics of the system.
Innovation grant from Google
Journalism++ Stockholm together with IDA PhD student Måns Magnusson gets an
innovation grant from Google News Initiative to develop Marple, a highly
automated news service for finding especially local stories in public data.
Prof. Erik Sandewall will be given the IJCAI 2016 Distinguished Service Award
The IJCAI Distinguished Service Award was established in 1979 by the IJCAI
Trustees to honor senior scientists in AI for contributions and service to the
field during their careers. Previous recipients have been Bernard Meltzer
(1979), Arthur Samuel (1983), Donald Walker (1989), Woodrow Bledsoe (1991),
Daniel G. Bobrow (1993), Wolfgang Bibel (1999), Barbara Grosz (2001), Alan
Bundy (2003), Raj Reddy (2005), Ronald J. Brachman (2007), Luigia Carlucci
Aiello (2009), Raymond C. Perrault (2011), Wolfgang Wahlster (2013) and Anthony
G. Cohn (2015).
At the 25th International Joint Conference on Artificial Intelligence, IJCAI-16, the Donald E. Walker Distinguished Service Award will be given to Erik Sandewall, Professor of Computer Science (retired) at the Department of Computer and Information Science at Linköping University. Professor Sandewall is recognized for his substantial contributions, as well as his extensive service to the field of Artificial Intelligence throughout his career. He is one of the founders of IJCAI and he served as the Editor-in-Chief of the Artificial Intelligence Journal for many years and made significant contributions to the success of the journal and to the wider dissemination of AI into the scientific community.
Analysis, Design, and Optimization of Embedded Control Systems
Today, many embedded or cyber-physical systems, e.g. in the automotive domain,
comprise several control applications, sharing the same platform. It is well
known that such resource sharing leads to complex temporal behaviors that
degrades the quality of control, and more importantly, may even jeopardize
stability in the worst case, if not properly taken into account.
In his thesis, Amir Aminifar considers embedded control or cyber-physical systems, where several control applications share the same processing unit. The focus is on the control-scheduling co-design problem, where the controller and scheduling parameters are jointly optimized. The fundamental difference between control applications and traditional embedded applications motivates the need for novel methodologies for the design and optimization of embedded control systems.
New technique for tropical cyclone forecasting
Chandan Roy's PhD thesis proposes a technique based on artificial neural
networks and open satellite data to get earlier and more accurate forecasts for
cyclone warning systems, especially in countries with less economic and
technical means like Bangladesh.
Design of Secure Real-Time Embedded Systems
Real-time embedded systems have several design time constraints such as timing,
resource efficieny and performance. Security has not been an important design
concern earlier, and has been overlooked in the past. Ke Jiang's PhD thesis at
IDA approaches the design of secure embedded systems with a focus on
communication confidentiality and side-channel attack resistance.
Several techniques are presented in this thesis for designing secure real-time embedded systems, including hardware/software co-design techniques for communication confidentiality on distributed platforms, a global framework for secure multi-mode real-time systems, and a scheduling policy for thwarting differential power analysis attacks.
All the proposed solutions have been extensively evaluated experimentally, including two real-life case studies, which demonstrate the efficiency of the presented techniques.
New algebraic methods to study the complexity of constraint satisfaction problems
Constraint satisfaction problems are an important type of computational
problems with many applications within for example scheduling and optimisation
problems. The so-called algebraic approach has for the past 20 years had
tremendous success in identifying constraint satisfaction problems solvable in
polynomial time. Since efficient algorithms exists for such problems, they are
often said to be tractable. However, this algebraic approach cannot be used to
study the worst-case time complexity for the problems that are not believed to
be tractable, even though these problems can vary substantially in complexity.
To study the difference in complexity between hard constraint satisfaction problems of this kind, Victor Lagerkvist in his PhD thesis on IDA proposes an algebraic framework based on partial functions. This framework is then used to study the complexity for many well-known variants of the constraint satisfaction problem. The complexity for these problems is then related to a complexity theoretical conjecture, the exponential-time hypothesis, which states that there is a sharp limit to how much it is possible to improve exponential-time algorithms for constraint satisfaction problems.
Page responsible: Webmaster