In the past ten years, boosting has become a major field of machine learning and classification. This paper brings contributions to its theory and algorithms. We first unify a well-known top-down decision tree induction algorithm due to [Kearns and Mansour, 1999], and discrete AdaBoost [Freund and Schapire, 1997], as two versions of a same higher-level boosting algorithm. It may be used as the ...

It is an already established fact through the investigations by Thomson (1), Saenz (2), Rist (3), Freund (4), Hensel (5), etc that the inoculation of a paraffin oil suspension of heat-killed tubercle bacilli produces in experimental animals a state of hypersensitivity much more intense and lasting than that given by the same amount of the killed bacilli in saline solution. We have confirmed thi...

We discuss two types of algorithms for selecting relevant examples that have been developed in the context of computation learning theory. The examples are selected out of a stream of examples that are generated independently at random. The rst two algorithms are the so-called \boosting" algorithms of Schapire Schapire, 1990] and Fre-und Freund, 1990], and the Query-by-Committee algorithm of Se...

The boosting algorithm AdaBoost de veloped by Freund and Schapire has ex hibited outstanding performance on sev eral benchmark problems when using C as the weak algorithm to be boosted Like other ensemble learning approaches AdaBoost constructs a composite hy pothesis by voting many individual hy potheses In practice the large amount of memory required to store these hypotheses can make ensembl...

We continue our previous research on cooperating distributed grammar systems (CDGS) and variants thereof as language acceptors 16]. Here, we classify the accepting capacity of CDGS working in the modes recently introduced by the authors together with Freund 14]. Moreover, we study (prescribed) teams as language accepting mechanisms. In this way, we solve an open problem from the area of accepti...

Motivated by a recent method of Freund [3], who introduced a quasi-minimal residual (QMR) version of the CGS algorithm, we propose a QMR variant of the Bi-CGSTAB algorithm of van der Vorst, which we call QMRCGSTAB for solving nonsymmetric linear systems. The motivation for both QMR variants is to obtain smoother convergence behavior of the underlying method. We illustrate this by numerical expe...

A family of generalized S-brane solutions with orthogonal intersection rules and n Ricci-flat factor spaces in the theory with several scalar fields, antisymmetric forms and multiple scalar potential is considered. Two subclasses of solutions with power-law and exponential behaviour of scale factors are singled out. These subclasses contain sub-families of solutions with accelerated expansion o...

partially fed (4 -5 days) h .a. anatolicum (iranian isolate) female adult ticks were used for preparation of antigens of midgut which are, the gut supernate antigen (gs), gut membrane antigen (gm), triton extracted gut membrane antigen ( tx-gm ) and sodium dodecyl sulphate treated triton extracted gut membrane antigen (sds-tx-gm). the new zealand white rabbits were immunized with gs, gm, tx- gm...

Treatment for sexual offenders has changed drastically over the past half-century (Laws and Marshall, 2003) . Even though earlier non-behavioral treatment approaches were important in establishing that child sex offenders could be engaged in treatment, John B . Watson and Alfred Kinsey were prominent in the development of behavioral treatment approaches in the early to mid-20th century. Kurt Fr...

Variable dimension algorithms are a class of algorithms for computation of fixed points. They normally start at a single point and generate a path of simplices of varying dimension until a simplex that contains an approximation of a fixed point is found. This thesis analyzes, compares, and contrasts five duality models for variable dimension fixed point algorithms, namely, primal-dual subdivide...