This paper describes an efficient sampler for synchronous grammar induction under a nonparametric Bayesian prior. Inspired by ideas from slice sampling, our sampler is able to draw samples from the posterior distributions of models for which the standard dynamic programing based sampler proves intractable on non-trivial corpora. We compare our sampler to a previously proposed Gibbs sampler and ...

This lecture explores the application of pixon-based methods to the reconstruction of complexly encoded images. The special problems sometimes associated with complexly encoded data sets, i.e. an extremely complex and convoluted solution space, are discussed, and numerical methods for dealing with these problems, e.g. simulated annealing, genetic algorithms, and evolutionary programing, are pre...

An optimization-based approach to fault diagnosis for nonlinear stochastic dynamic models is developed. An optimal diagnosis problem is formulated according to a receding-horizon strategy. This approach leads to a functional optimization problem (also called ‘infinite optimization problem’), whose admissible solutions belong to a function space. As in such a context, the tools from mathematical...

Isaac is a programing language for geometric reasoning intended for controlling mobile robots, currently under development at NMSU. Due to the application area, handling input and output (in particular, sensor input and actuator output) in a manner consistent with the language as a whole is particularly important in this language. In this paper, we discuss the mechanisms employed by Isaac to pr...

[ABSTRACT] The paper presents a data–driven execution model, CHEM, for a temporal logic programing language, Chronolog. An intermediate virtual machine is proposed, which is granulated at clause argument level to exploit argument parallelism through unification. Context–parallelism, inherent in temporal logic programs, is exploited through dynamic tagging approach typically used in dataflow com...

We use the duality of linear programing to obtain exact formulas of competitive ratio for the semi-online preemptive scheduling on three and four machines. We use the linear programs from [5]. Namely we consider the online scheduling and the semi-online scheduling with known sum of processing times. We solve the linear programs symbolically by examining all basic solution candidates. All soluti...

In the current state-of-art beat tracker, it is common that a specific algorithm do the best on particular kind of audio. For example, the algorithm based on two-fold dynamic programing have sequential 1 st place on time-varying-tempo MAZ dataset of MIREX audio beat tracking. This study proposed a tempogram-sensing-vector (TSV) method to be adaptive to time-varying-tempo and stable-tempo excerp...

The target approach to reactor synthesis tries to determine an upper bound for the performance of a chemical reactor network as measured by its objective function (such as yield or reactor volume). This leads to an optimal control formulation in which the residence time distribution and the micromixing function are the control variables. The optimal control problem is solved using a gradient ba...

In this paper an exact penalty line search is introduced for solving constrained nonlinear programing. By this method it is possible to solve some problems with combining penalty method and SQP method, and in any iteration for producing new stepone can uses linear search and solves unconstrained optimization by exact penalty function. In this method a linear programming subproblem with trust ar...