论文信息 - A Probabilistic Approach to Identifiability

A Probabilistic Approach to Identifiability

The structural properties of the probabilistic coupling between the unknown parameters of a dynamical system and the observation process are studied. Basic tools for the analysis are the dual concepts of sufficient and observable statistics introduced in the paper. If the observation noise acts in an additive fashion then the transition probability admits a density displaying an exponential family — like structure.

Giorgio Picci | G. Picci

[1] V. Popov. Some properties of the control systems with irreducible matrix — Transfer functions , 1970 .

[2] J. Rissanen. Basis of invariants and canonical forms for linear dynamic systems , 1974, Autom..

[3] L. J. Savage,et al. Application of the Radon-Nikodym Theorem to the Theory of Sufficient Statistics , 1949 .

[4] R. Brockett. System Theory on Group Manifolds and Coset Spaces , 1972 .

[5] R. R. Bahadur. Sufficiency and Statistical Decision Functions , 1954 .

[6] G. Kallianpur,et al. Arbitrary system process with additive white noise observation errors , 1968 .

[7] Thomas Kailath,et al. ABSOLUTE CONTINUITY AND RADON-NIKODYM DERIVATIVES FOR CERTAIN MEASURES RELATIVE TO WIENER MEASURE' , 1971 .