PL
 |
EN
Preferencje help
Polish version English version Język
Widoczny [Schowaj] Abstrakt
Liczba wyników

Znaleziono wyników: 11

Liczba wyników na stronie
first rewind previous Strona / 1 next fast forward last
Wyniki wyszukiwania
Wyszukiwano:
w słowach kluczowych:  multiple criteria analysis
help Sortuj według:

help Ogranicz wyniki do:
first rewind previous Strona / 1 next fast forward last
1
Content available A multi-criteria approach for selecting an explanation from the set of counterfactuals produced by an ensemble of explainers
Stepka Ignacy, Lango Mateusz, Stefanowski Jerzy
International Journal of Applied Mathematics and Computer Science
|
2024
|
Vol. 34, no. 1
119--133
EN
Counterfactuals are widely used to explain ML model predictions by providing alternative scenarios for obtaining more desired predictions. They can be generated by a variety of methods that optimize various, sometimes conflicting, quality measures and produce quite different solutions. However, choosing the most appropriate explanation method and one of the generated counterfactuals is not an easy task. Instead of forcing the user to test many different explanation methods and analysing conflicting solutions, in this paper we propose to use a multi-stage ensemble approach that will select a single counterfactual based on the multiple-criteria analysis. It offers a compromise solution that scores well on several popular quality measures. This approach exploits the dominance relation and the ideal point decision aid method, which selects one counterfactual from the Pareto front. The conducted experiments demonstrate that the proposed approach generates fully actionable counterfactuals with attractive compromise values of the quality measures considered.
2
Content available remote Applying Data Envelopment Analysis Principle in Ordinal Multi Criteria Decision Analysis
Kress Moshe
Foundations of Computing and Decision Sciences
|
2021
|
Vol. 46, No. 2
147--157
EN
We consider a multicriteria decision analysis (MCDA) problem where importance of criteria, and evaluations of alternatives with respect to the criteria, are expressed on a qualitative ordinal scale. Using the extreme-point principle of Data Envelopment Analysis (DEA), we develop a two-parameter method for obtaining overall ratings of the alternatives when preferences and evaluations are made on an ordinal scale. We assume no parametric setup other than the two parameters that reflect minimum intensities of discriminating among rank positions: one parameter for the alternatives’ ranking and one for the criteria ranking. These parameters are bounded by the ordinal input data, and they imply a universal tie among the alternatives when both parameters are selected to be zero. We describe the model, discuss its theoretical underpinning, and demonstrate its application.
3
Content available Modelling uncertainties in multi-criteria decision making using distance measure and topsis for hesitant fuzzy sets
Beg I., Rashid T.
Journal of Artificial Intelligence and Soft Computing Research
|
2017
|
Vol. 7, No. 2
103--109
EN
A notion for distance between hesitant fuzzy data is given. Using this new distance notion, we propose the technique for order preference by similarity to ideal solution for hesitant fuzzy sets and a new approach in modelling uncertainties. An illustrative example is constructed to show the feasibility and practicality of the new method.
4
Content available Three step procedure for a multiple criteria problem of project portfolio scheduling
Krzeszowska B.
Operations Research and Decisions
|
2013
|
Vol. 23, No. 4
55--74
EN
The project portfolio scheduling problem as a multiple criteria decision making problem and a three step procedure to solve this problem have been presnted. In the first step, the problem was described by a multiple criteria mathematical model. Three criteria have been considered: minimization of the sum of penalties for projects delays, minimization of resource overuse and NPV maximization. In the second step, non-dominated solutions were identified by using an elitist evolutionary algorithm for multiple optimization. In the third step, an interactive procedure has been applied to choose the final solution. An example of a portfolio of IT projects was used for computations.
5
Content available remote Integrating imperfection of information into the PROMETHEE multicriteria decision aid methods: A general framework
Ben Amor S., Mareschal B.
Foundations of Computing and Decision Sciences
|
2012
|
Vol. 37, No. 1
9-8
EN
Multicriteria decision aid methods are used to analyze decision problems including a series of alternative decisions evaluated on several criteria. They most often assume that perfect information is available with respect to the evaluation of the alternative decisions. However, in practice, imprecision, uncertainty or indetermination are often present at least for some criteria. This is a limit of most multicriteria methods. In particular the PROMETHEE methods do not allow directly for taking into account this kind of imperfection of information. We show how a general framework can be adapted to PROMETHEE and can be used in order to integrate different imperfect information models such as a.o. probabilities, fuzzy logic or possibility theory. An important characteristic of the proposed approach is that it makes it possible to use different models for different criteria in the same decision problem.
6
Content available remote Evolutionary multi-objective optimization for inferring outranking model’s parameters under scarce reference information and effects of reinforced preference
Fernandez E., Navarro J., Mazcorro G.
Foundations of Computing and Decision Sciences
|
2012
|
Vol. 37, No. 3
163--184
EN
Methods based on fuzzy outranking relations constitute one of the main approaches to multiple criteria decision problems. The use of ELECTRE methods require the elicitation of a large number of parameters (weights and different thresholds) but direct eliciting is often a demanding task for the decision-maker (DM). For handling intensity-of-preference effects on concordance levels, a generalized concordance model was proposed by Roy and Slowinski which is more complex than previous outranking models. In this paper, an evolutionary multi-objective-based indirect elicitation of the complete ELECTRE III model-parameter set is proposed. The evolutionary multi-objective inference method is successfully extended to inferring reinforced-preference model parameters. Wide experimental evidence is provided to support the proposal, which performs well even working on small size reference sets.
7
Content available Incorporating Customer Preference Information into the Forecasting of Service Sales
Rzepakowski P.
Journal of Telecommunications and Information Technology
|
2009
|
nr 3
50-58
EN
Customers change their preferences while getting more familiar with services or being motivated to change their buying habits. Different sources of motivation induce customers to change their behavior: an advertisement, a leader in a reference group, satisfaction from services usage and other experiences, but usually those reasons are unknown. Nevertheless, people vary in susceptibility to suggestions and innovations, and also in preference structure change dynamics. Historical information about the preference structure gives additional information about uncertainty in forecasting activity. In this work the conjoint analysis method was used to find customer preference structure and to improve a prediction accuracy of telecommunication services usage. The results have shown that prediction accuracy increases about by one percent point, what results in a 20 percent increase after using proposed algorithm modification.
8
Content available Supporting telecommunication product sales by conjoint analysis
Rzepakowski P.
Journal of Telecommunications and Information Technology
|
2008
|
nr 3
28-34
EN
Conjoint analysis is widely used as a marketing research technique to study consumers' product preferences and simulate customer choices. It is used in designing new products, changing or repositioning existing products, evaluating the effect of price on purchase intent, and simulating market share. In this work the possibility of conjoint analysis usage in telecommunication filed is analyzed. It is used to find optimal products which could be recommended to telecommunication customers. First, a decision problem is defined. Next, the conjoint analysis method and its connections with ANOVA as well as regression techniques are presented. After that, different utility functions that represent preferences for voice, SMS, MMS and other net services usage are formulated and compared. Parameters of the proposed conjoint measures are determined by regression methods running on behavioral data, represented by artificially generated call data records. Finally, users are split in homogenous groups by segmentation techniques applied to net service utilities derived from conjoint analysis. Within those groups statistical analyses are performed to create product recommendations. The results have shown that conjoint analysis can be successfully applied by telecommunication operators in the customer preference identification process. However, further analysis should be done on real data, other data sources for customer preference identification should be explored as well.
9
Content available On the performance analysis of a heuristic approach dedicated to a multiobjective dynamic routing model
Craveirinha J., Martins L., Climaco J. N., Jorge L.
Journal of Telecommunications and Information Technology
|
2005
|
nr 3
14-23
EN
In previous works the features and a complete formulation for circuit-switched networks of a multiple objective dynamic routing method (MODR) of periodic state dependent routing type were presented. The aim of the model is to resolve a very complex network bi-objective dynamic routing problem, by recurring to a heuristic for synchronous path selection enabling to obtain a good compromise solution in terms of two network performance measures. In this paper we present a study on the performance of variants of the MODR heuristic of synchronous path selection by using relaxations of the values previously calculated for the two network objective functions. This study permitted the development of an improved version of the initial heuristic. Also a comparison of the analytical values of the network objective functions obtained with selected variants of the initial heuristic with the corresponding results from a known reference method, the real time network routing (RTNR) method, given by a discrete-event simulator for single-service networks, is presented.
10
Sustainable urban development and web-based multiple criteria analysis
Zavadskas E. K., Kaklauskas A., Saparauskas J.
Foundations of Civil and Environmental Engineering
|
2005
|
No. 6
217-226
EN
The analysis of information, expert and decision support systems used in sustainable urban development that were developed by researchers from various countries assisted the authors in the creation of their own Sustainable Urban Development Decision Support System (SUD-DSS). SUD-DSS differs from other systems in the use of new multiple criteria analysis methods as were developed by the authors. The database of a sustainable urban development was developed providing a comprehensive assessment of alternative versions from the economic, political, legal, social, management, technical, technological and other points of view. Based on the above complex databases, the developed SUD-DSS enables the user to analyse alternatives quantitatively (i.e. a system and subsystems of criteria, units of measure, values and weights) and conceptually (i.e. the text, formula, schemes, graphs, diagrams and videotapes).
PL
Strategia harmonijnego rozwoju konkretnego miasta nie może być ślepo naśladowana. Może ona jedynie być dostosowywana do warunków ekonomicznych, społecznych, politycznych, prawny i innych. Nie istnieje jakaś uniwersalna strategia harmonijnego rozwoju miasta pasująca dla jakiegokolwiek społeczeństwa lub kraju. Analiza systemów wspierania decyzji informacyjnych, ekspertowych utworzonych przez naukowców z różnych krajów pobudziła autorów do stworzenia swojego systemu wspierania decyzji harmonijnego rozwoju miasta (SWDHRM). System ten od innych systemów różni się tym, że są w nim stosowane nowe, stworzone przez autorów metody analizy wielowymiarowej. Także została stworzona ba/a danych mogąca służyć do szczegółowej oceny perspektyw ekonomicznych, politycznych, prawnych, społecznych, zarządczych, technologicznych oraz ich wersji. Przy użyciu wspomnianej bazy danych w utworzonym SWDHRM można przeprowadzać ilościową i jakościową analizę kolejnych wersji.
11
Content available remote Interactive approach in multicriteria analysis based on stochastic dominance
Nowak M.
Control and Cybernetics
|
2004
|
Vol. 33, no 3
463-476
EN
The paper considers a discrete stochastic multicriteria problem. This problem can be denned by a finite set of actions A, a set of attributes X and a set of evaluations E. It is assumed that the performance probability distributions for each action on each attribute are known. A new procedure for such a problem is proposed. It is based on two concepts: stochastic dominance and interactive approach. Stochastic dominance is employed for comparing evaluations of actions with respect, to attributes. The STEM methodology is employed in the dialogue procedure between decision maker and decision model. In each step a candidate action a_i is generated. The decision maker examines evaluations of a_i, with respect to attributes and selects the one that satisfies him/her. Then the decision maker defines the limit of concessions, which can be made on average evaluations with respect to this attribute. The procedure continues until a satisfactory action is found.
first rewind previous Strona / 1 next fast forward last
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.