Competitive training in hidden Markov models

The use of hidden Markov models is placed in a connectionist framework, and an alternative approach to improving their ability to discriminate between classes is described. Using a network style of training, a measure of discrimination based on the a posteriori probability of state occupation is proposed, and the theory for its optimization using error back-propagation and gradient ascent is presented. The method is shown to be numerically well behaved, and results are presented which demonstrate that when using a simple threshold test on the probability of state occupation, the proposed optimization scheme leads to improved recognition performance.

Graphene-driven revolutions in ICT and beyond

This session described the FET Flagship Pilot on graphene and related two-dimensional materials. The flagship targets a revolution in information and communication technology, with impacts reaching into other areas of the society. The session featured four talks on the scientific and technological potential and open research challenges within the scope of the proposed flagship, industrial view on possibilities and challenges posed by graphene and related materials, and presentation on the implementation and structure of the flagship pilot. © Selection and peer-review under responsibility of FET11 conference organizers and published by Elsevier B.V.