We address the problem that different users have different lexical knowledge about problem domains, so that automated dialogue systems need to adapt their generation choices online to the users’ domain knowledge as it encounters them. We approach this problem using policy learning in Markov Decision Processes (MDP). In contrast to related work we propose a new statistical user model which incor...

Most studies on adaptive partitioning policies for scheduling parallel jobs on distributed memory parallel computers ignore the constraints imposed by the memory requirements of the jobs. In this paper , we rst show that these constraints can have a negative impact on the performance of adaptive partitioning policies. We then evaluate the performance of adaptive partitioning in a system where t...

Most studies on adaptive partitioning policies for scheduling parallel jobs on distributed memory parallel computers ignore the constraints imposed by the memory requirements of the jobs. In this paper , we rst show that these constraints can have a negative impact on the performance of adaptive partitioning policies. We then evaluate the performance of adaptive partitioning in a system where t...

p Ž . 0 Ž . Ž . Ž . V u under any policy p in C is equal to nm* u y M u log n q o log n , as n R Ž . Ž . Ž . n a `, where m* u is the largest population mean and M u is a constant. ii Policies in C are asymptotically optimal within a larger class C of ‘‘uniformly R UF 0 p Ž Ž . Ž .. fast convergent’’ policies in the sense that limna ` nm* u y V u r n Ž Ž . pŽ .. Ž . nm* u y V u F 1, for any p g...

Public policies must be devised in spite of profound uncertainties about the future. When there are many plausible scenarios for the future, it may well be impossible to construct any single static policy that will perform well in all of them. It is likely, however, that the uncertainties that confront planners will be resolved over the course of time by new information. Thus, policies should b...

This work supported in part by the National Science Foundation under grant ECS-9904378. Abstract Adaptive critic based approximate dynamic programming techniques are gradient based methods for finding optimal policies for multi-stage decision processes. We believe adaptive critic methods are now developed to the point that they can be applied to the full spectrum of decision and control problem...

Schumpeter chastised Ricardo for his alleged "vice" - the so-called "Ricardian Vice" - of drawing far reaching policy conclusions from utterly simplistic models, which, moreover, were underdetermined. The paper first argues that Schumpeter saw Ricardo's approach to the theory of value and distribution through a marginalist lens and therefore arrived at a distorted picture of the latter. Several...

Just as for people, a new year is a good time for a journal to stop, look back, and start a new beginning, hoping for a good season ahead. This is what we have been doing each year in the Croatian Medical Journal for the last twenty years. We will probably never stop doing this, even when we become an august journal, because the world of scientific publishing rapidly changes. This year, we pres...

The journals of the International Union of Crystallography have grown in size and number over the past 60 years to match developments in scientific practice and technique. High quality of publication has always been at the forefront of editorial policy and ways in which this has been achieved are described. In particular, the development of standard exchange and archive formats for crystallogra...

We introduce here the inaugural issue of the new scientific journal Translation. The overarching aim of this endeavor is to establish a new forum for a broad spectrum of research in the area of protein synthesis in living systems ranging from structural biochemical, evolutionary and regulatory aspects of translation to the fundamental questions related to post-translational control of somatic p...