The workshop on Symbolic-Numeric methods for Reasoning about CPS and IoT (SNR) focuses on the combination of symbolic and numeric methods for reasoning about Cyber-Physical Systems and the Internet of Things to facilitate model identification, specification, verification, and control synthesis problems for these systems. The synergy between symbolic and numerical approaches is fruitful for two main reasons:
- Symbolic methods that operate on exact and discrete representations of systems, the set of reachable states, the distribution of model parameters or the possible gains for controller parameters.
- Numeric methods that operate on various forms of numerical approximations and continuous transformations of the systems, as developed in the area of continuous dynamical systems and control theory.
Such synergies are already seen in areas such as reachability analysis (symbolic representation of reachable states versus numerical integration), uncertainty reasoning (eg., Rao-Blackwellization), machine learning (eg., learning models through stochastic gradient descent versus symbolic reasoning over the function represented by the network to prove properties) and decision procedures (eg., symbolic SAT solvers versus numerical convex optimization solvers).
The SNR workshop aims to catalyze work on the interface of symbolic and numeric methods for verification, synthesis and identification problems for CPS and IoT. The scope of the workshop includes, but is not restricted to, the following topics:
- Verification, parameter identification and control synthesis for hybrid systems.
- Probabilistic inference and reachability for stochastic hybrid systems.
- Symbolic and Numerical integration techniques.
- Symbolic-Numeric decision procedures.
- Emerging applications to safe autonomous systems.
We encourage submissions of papers in the following two specific areas:
- Symbolic and Numeric Methods for Artificial Intelligence.
Learning algorithms are at the core of many engineering applications including robotics and autonomous vehicles.
We invite research papers on verification of models used in machine learning and autonomous CPS.
In particular, recent advances in autonomous cars require addressing challenging questions around their safety and reliability.
- Verification and Synthesis of Stochastic Models.
Autonomous systems operate in uncertain environments. Thus, it is essential to reason about the effect of uncertainty. We invite research papers on symbolic and numerical techniques for formal synthesis of stochastic systems.