This paper proposes a method to solve problem that comes with imbalanced training sets which is often seen in the practical applications. We modied Self Growing Neural Network CombNET-II to deal with the imbalanced condition. This model is then applied to practical application which was launched in '99 Fog Forecasting Contest sponsored by Neurocomputing Technical Group of IEICE, Japan. In this ...

Large VC-dimension classiers can learn dicult tasks, but are usually impractical because they generalize well only if they are trained with huge quantities of data. In this paper we show that even very high-order polynomial classiers can be trained with a small amount of training data and yet generalize better than classiers with a smaller VC-dimension. This is achieved with a maximum margin al...

The representation of complex spatial domains in conventional databases suuers from fragmented representation of object structure , lack of instance-level spatial relationships, and the generation of large combinatoric search spaces in query analysis. The deductive capabilities provided by a deductive database ooer some assistance in solving these problems, in particular by enabling spatial rea...

A wide variety of problems in control system theory fall within the class of parameterized Linear Matrix Inequalities (LMIs), that is, LMIs whose coecients are functions of a parameter conned to a compact set. However, in contrast to LMIs, param-eterized LMI (PLMIs) feasibility problems involve innitely many LMIs hence are very hard to solve. In this paper, we propose several eective relaxation...

the bi-objective warehouse problem in a crisp environment is often not eective in dealing with the imprecision or vagueness in the values of the problem parameters. to deal with such situations, several researchers have proposed that the parameters be represented as fuzzy numbers. we describe a new algorithm for fuzzy bi-objective warehouse problem using a ranking function followed by an applic...

This paper is about concept maps and how they can assist in the learning ofconcepts of mathematics. First the paper presents the theoretical backgroundand working denitions for concept maps. Then this study examines the impactof using concept maps in learning of fractions. Results of this study indicatedthat using this strategy was eective in learning of fractions for fourth-gradestudents. This...

The bi-objective warehouse problem in a crisp environment is often not eective in dealing with the imprecision or vagueness in the values of the problem parameters. To deal with such situations, several researchers have proposed that the parameters be represented as fuzzy numbers. We describe a new algorithm for fuzzy bi-objective warehouse problem using a ranking function followed by an applic...

In this article, a mathematical model describing the growth orterminating myelogenous leukemia blood cancer's cells against naive T-celland eective T-cell population of body, presented by fractional dierentialequations. We use this model to analyze the stability of the dynamics, whichoccur in the local interaction of eector-immune cell and tumor cells. Wewill also investigate the optimal contro...

In this article we found the solution of fuzzy linear controlled systemwith fuzzy initial conditions by using -cuts and presentation of numbersin a more compact form by moving to the eld of complex numbers. Next, afuzzy optimal control problem for a fuzzy system is considered to optimize theexpected value of a fuzzy objective function. Based on Pontryagin MaximumPrinciple, a constructive equati...

In the eld of nonlinear programming (in continuous variables) convex analysis [21, 22] plays a pivotal role both in theory and in practice. An analogous theory for discrete optimization (nonlinear integer programming), called \discrete convex analysis" [18, 17], is developed for L-convex and M-convex functions by adapting the ideas in convex analysis and generalizing the results in matroid theo...