We present a new approximation algorithm based on an exact representation of the state space S, using decision diagrams, and of the transition rate matrix R, using Kronecker algeb...
Andrew S. Miner, Gianfranco Ciardo, Susanna Donate...
This paper describes recent advances in hidden Markov model (HMM) based OCR for machine-printed Arabic documents. A combination of scriptindependent and script-specific techniques...
This paper presents a new prediction model for predicting when an online customer leaves a current page and which next Web page the customer will visit. The model can forecast the ...
In this paper, we propose a new recognition method of lung nodules from X-ray CT images using 3D Markov random field(MRF) models. Pathological shadow candidates are detected by a...
We describe a new tagging model where the states of a hidden Markov model (HMM) estimated by unsupervised learning are incorporated as the features in a maximum entropy model. Our...