Variational Bayes (VB), also known as independent mean-field approximation, has become a popular method for Bayesian network inference in recent years. Its application is vast, e.g. in neural network, compressed sensing, clustering, etc. to name just a few. In this paper, the independence constraint in VB will be relaxed to a conditional constraint class, called copula in statistics. Since a jo...

Weakly relational numeric domains express restricted classes of linear inequalities that strike a balance between what can be described and what can be efficiently computed. Such domains often restrict their attention of TVPI constraints which are systems of constraints where each constraint involves, at most, two variables. This technical report addresses the problem of deriving an incremental...

Many AI tasks can be formalized as constraint satisfaction problems (CSPs), which involve finding values for variables subject to a set of constraints. While solving a CSP is an NP-complete task in general, it is believed that efficiency can be significantly improved by exploiting the characteristics of the problem. In this paper, we present a solution synthesis algorithm called ω-CDGT which is...

The varestraint cracking test was carried out to assess the hot cracking tendency of alloy 800 at different welding conditions and augmented strain levels. The test results revealed that the hot cracking susceptibility of alloy 800 increases with increasing the heat input and augmented strain levels. This emphasizes the role of constraint either internally induced or externally applied to the w...

We consider a class of constrained optimization problems where the objective function is a sum of a smooth function and a nonconvex non-Lipschitz function. Many problems in sparse portfolio selection, edge preserving image restoration and signal processing can be modelled in this form. First we propose the concept of the Karush-Kuhn-Tucker (KKT) stationary condition for the non-Lipschitz proble...

In this paper, the problem of finding the set of all real solutions to a system of n nonlinear general form algebraic equations contained in a given n-dimensional box (the global analysis problem) is considered. A new iterative interval method for solving the global analysis problem is suggested. It is based on the following techniques: (i) transformation of the original system into an augmente...

Traditionally, constraint satisfaction has been applied in closedworld scenarios, where all choices and constraints are known from the beginning and fixed. With the Internet, many of the traditional CSP applications in resource allocation, scheduling and planning pose themselves in open-world settings, where choices and constraints are to be discovered from different servers in a network. We ex...

The use of visual languages in computing is varied, ranging from system modelling to the display of data analyzed in computation processes. A prominent example of a visual notation is the Unified Modelling Language (UML), designed for use by software engineers. Constraint diagrams were proposed as an alternative to the UML’s Object Constraint Language. Spider diagrams form a fragment of constra...

In this paper we present a hybrid method, named Quasi Local Search, which combines Simulated Annealing augmented with a type of Tabu List to guide the search globally with Constraint Programming to search locally the optimal solution. The method can also be seen as an integration framework, once the Local Search module is independent of the Constraint Programming one and either can be worked in...

We present a row-by-row (RBR) method for solving semidefinite programming (SDP) problem based on solving a sequence of problems obtained by restricting the n-dimensional positive semidefinite constraint on the matrix X. By fixing any (n − 1)-dimensional principal submatrix of X and using its (generalized) Schur complement, the positive semidefinite constraint is reduced to a simple second-order...