We present a new estimation principle for parameterized statistical models. The idea is to perform nonlinear logistic regression to discriminate between the observed data and some...
In this paper we extend the control methodology based on Extended Markov Tracking (EMT) by providing the control algorithm with capabilities to calibrate and even partially recons...
Consider a hidden Markov chain obtained as the observation process of an ordinary Markov chain corrupted by noise. Zuk, et. al. [13, 14] showed how, in principle, one can explicit...
The paper introduces a new detectability concept for continuous-time Markov jump linear systems with finite Markov space that generalizes previous concepts found in the literature....
This paper presents a hybrid approach for named entity (NE) tagging which combines Maximum Entropy Model (MaxEnt), Hidden Markov Model (HMM) and handcrafted grammatical rules. Eac...