Radha Jagadeesan - Continuous Markov processes --- Approximations and metrics

Partial Labeled Markov processes provide a foundation for interacting probabilistic systems with continuous state spaces. Such systems arise in the analysis of stochastic physical systems or stochastic hybrid systems.

We develop contextual reasoning principles for reasoning about robust notions of "approximate" equivalence between concurrent interacting probabilistic systems. The key technical tools in this development are:

  1. An approximation theory of continuous systems by a countable family of finite-state probabilistic systems.
  2. A move from traditional logic to real valued functions as formulas (as advocated by Kozen)

(Joint work with Josee Desharnais, Vineet Gupta and Prakash Panangaden)