Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: ranking lists

Sortuj według:

Ogranicz wyniki do:

Compromise in scheduling objects procedures basing on ranking lists

Piech H., Skibińska W.

Prace Naukowe Instytutu Matematyki i Informatyki Politechniki Częstochowskiej

2010

Vol. 9, nr 1

177-186

In our work possibility of support of ranking objects (tasks) is analyzed on base of group of lists. We can get these lists both from experts or with help of approximating and simple (according to complexity) algorithms. To support analyze we can use elements of neighborhood theory, preferential models, and rough sets theory. This supporting process is used for creation final list of tasks sequence. Usually, these problems are connected with distribution, classification, prediction, strategy of games as well as compromise searching operations. The utilization preference and domination models permits to crisp inferences and to force the chronological location of object. In some situations we have deal with dynamic character of filling lists resulting from continuous tasks succeeding and continuous their assigning to executive elements. The utilization the theory of neighborhood permits to locate objects in range of compromised solutions consist in closing to dominating proposal group. Main task for us is find the best compromise in aspect to final objects location. We want to defined advantages and drawback of methods basing on mention theories and analyze possibilities of their cooperation or mutual completions.

Combing up method for creating final objects location on base of ranking lists

Piech H., Chłopaś T.

Prace Naukowe Instytutu Matematyki i Informatyki Politechniki Częstochowskiej

2009

Vol. 8, nr 1

161-171

Multi-attribute classification (ordering) problem concerns assignment of objects (tasks) to some predefined and preference-ordered decision classes [1, 2]. In considered situation objects are described by a finite sets of locations. Disposing several objects ranking lists it is possible exploit information about distribution. It was assumed that distribution information base on so called neighborhoods structures. These structures can approve or disapprove judgments unanimity prepared by experts (agents, responders, algorithms). Disposing distribution characteristics we define set of criteria. So, it is challenging problem which can be solved on base of preference theory. Reaching preference approach by neighborhood elements we exploit addition information, which help compromise achievement. Proposed method of objects ordering based of considering preference relations during process of lists combining up along their positions. Such approach enlarge scale of divergences in these relations.