Estimation and Reconstruction for Stochastic Processes and Deterministic Functions.
Abstract
Method for statistical estimation of parameters of partially observed stochastic processes and for minimum mean squared error reconstruction of unobserved portions of sample paths (state estimation) were developed. Some of these methods apply to models of random distributions of particles in space or events in time; others apply to Markov processes. Statistical estimators are asymptotically exact even though certain of the unknown parameters are infinite-dimensional. For several classes of processes the problem of simultaneously performing parameter estimation and state estimation was solved. Refined techniques for reconstructing a deterministic signal from hard-limited data were devised. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 14, 1983
- Accession Number
- ADA136571
Entities
People
- A. F. Karr
Organizations
- Johns Hopkins University