Showing papers on "Pairwise comparison published in 1995"
TL;DR: A goal programming method is presented that has, as its objective, the generation of the priority vector whose associated comparison values are, on average, the closest to the pairwise comparison information provided by the evaluator.
Abstract: The generation of priority vectors from pairwise comparison information is an integral part of the Analytic Hierarchy Process (AHP). Traditionally, either the right eigenvector method or the logarithmic least squares method have been used. In this paper, a goal programming method (GPM) is presented that has, as its objective, the generation of the priority vector whose associated comparison values are, on average, the closest to the pairwise comparison information provided by the evaluator. The GPM possesses the properties of correctness in the consistent case, comparison order invariance, smoothness, and power invariance. Unlike other methods, it also possesses the additional property that the presence of a single outlier cannot prevent the identification of the correct priority vector. The GPM also has a pair of naturally meaningful consistency indicators that offer the opportunity for empowering the decision maker. The GPM is thus an attractive alternative to other proposed methods.
208 citations
TL;DR: This work simulates and analyzes the parameters of pairwise sequencing projects including template length, sequence read length, and total sequence redundancy and finds that pairwise strategies are effective with both small (cosmid) and large (megaYAC) targets and produce ordered sequence data with a high level of mapping completeness.
148 citations
TL;DR: In this article, the authors developed interactive approaches for the discrete alternative multiple criteria decision making problem, where the objective is to find the most preferred alternative of a decision maker for a DM assuming only that the DM has a monotonic utility function.
Abstract: In this paper we develop interactive approaches for the discrete alternative multiple criteria decision making problem. We develop an algorithm that finds the most preferred alternative of a decision maker DM assuming only that the DM has a monotonic utility function. The algorithm divides the criteria space into a number of smaller subspaces and then uses the ideal points of these subspaces to eliminate alternatives. We also develop a more efficient version of the algorithm for the more restrictive case of a monotonic quasiconcave utility function. We present favorable computational results in terms of the required number of pairwise comparisons for both versions of the algorithm. We then develop a general algorithm that first identifies the type of the DM's utility function and then employs the approach that is compatible with the identified utility function type. We also present computational results for the general algorithm.
76 citations
TL;DR: In this paper, the hyperbolic cosine unfolding model for direct responses of persons to individual stimuli is elabo rated in three ways: the parameter of the stimu lus, which reflects a region within which people located there are more likely to respond positively than negatively, is shown to be a property of the data and not arbitrary as first supposed.
Abstract: The hyperbolic cosine unfolding model for direct responses of persons to individual stimuli is elabo rated in three ways. First, the parameter of the stimu lus, which reflects a region within which people located there are more likely to respond positively than negatively, is shown to be a property of the data and not arbitrary as first supposed. Second, the model is used to construct a related model for pairwise preferences. This model, for which joint maximum likelihood estimates are derived, satisfies strong sto chastic transitivity. Third, the role of substantive theory in evaluating the fit between the data and the models, in which unique solutions for the estimates are not guaranteed, is explored by analyzing responses of one group of persons to a single set of stimuli obtained both as direct responses and pairwise preferences.
70 citations
TL;DR: In this paper, the authors show that requiring equilibrium to be pairwise proof (i.e., immune to bilateral deviations by the supplier and any customer) can imply non-existence of equilibrium in "normal" environments.
56 citations
TL;DR: In this article, a pseudo-likelihood method is used to estimate the parameters of the linear logistic test model. But the method is extended to estimate parameters of a linear model allowing the design matrix to vary between persons, and the pseudo likelihood estimates were comparable to conditional and marginal maximum likelihood estimates.
Abstract: Rasch model item parameters can be estimated consistently with a pseudo-likelihood method based on comparing responses to pairs of items irrespective of other items. The pseudo-likelihood method is comparable to Fischer's (1974) Minchi method. A simulation study found that the pseudo-likelihood estimates and their (estimated) standard errors were comparable to conditional and marginal maximum likelihood estimates. The method is extended to estimate parameters of the linear logistic test model allowing the design matrix to vary between persons.
52 citations
TL;DR: A new approach to cluster analysis of structures based on collective superpositions rather than pairwisesuperpositions is presented, which is fast and rigorous and illustrated by application to 21 structures derived from NMR experiments.
Abstract: A new approach to cluster analysis of structures based on collective superpositions rather than pairwise superpositions is presented. The method is fast and rigorous and is illustrated by application to 21 structures derived from NMR experiments. Source code, suitable for most laboratory machines, is available from the author, and a CCP4 version is in preparation.
42 citations
TL;DR: This model provides a mathematical explanation for the phenomena of intransitivity and inconsistency that sometimes appear in situations involving scoring on a subjective/qualitative domain and develops methods for generating consistent intervals on ratio scales.
40 citations
TL;DR: This paper extended the literature on the estimation of expected utility and non-expected utility preference functionals to a comparison of two different ways (pairwise choice and complete ranking) of experimentally obtaining data on such preferences.
Abstract: This paper extends the literature on the estimation of expected utility and non-expected-utility preference functionals (and the consequent exploration of the superiority of non-expected-utility over expected utility preference functionals) to a comparison of two different ways (pairwise choice and complete ranking) of experimentally obtaining data on such preferences. What is revealed is that the magnitude of the subject error is clearly conditional on the elicitation method used and, rather alarmingly, that the preference functional apparently employed by the subject may also be conditional on the elicitation method.
39 citations
TL;DR: It is shown how these swap operators can be used to compute bounds for Branch and Bound procedures, to prove dominance properties and to define a high-performance neighborhood for local search methods.
38 citations
TL;DR: Although the discussion is restricted to simple bilevel hierarchies, useful insight is provided into the methodologies in general and, in particular, the problem of inverse inconsistency associated with rank reversals.
Abstract: This paper discusses a method of comparing different prioritization techniques that entails mapping the input data from one system to another. The ranking vectors output by the methods should then be comparable. As an illustration, a method that uses pairwise comparison, e.g. the Analytic Hierarchy Process (AHP) of Saaty, itself viewed as a mapping, is compared with a method using cross impact analysis, the Edinburgh Approach. The problem centres on an appropriate inverse of the process mapping. Although the discussion is restricted to simple bilevel hierarchies, useful insight is provided into the methodologies in general and, in particular, the problem of inverse inconsistency associated with rank reversals. A way in which the theory of AHP might be incorporated into a Differential Hierarchy Process is suggested.
TL;DR: This paper defines camera break and proposes a new video partitioning algorithm, called NET Comparison (NC), which compares the pixels along predefined net lines, which processes images two to three times faster than the best existing approach.
Abstract: As video information proliferates, managing video sources becomes increasingly important. Automatic video partitioning is a prerequisite for organizing and indexing video sources. Several methods have been introduced to tackle this problem, e.g., pairwise and histogram comparisons. Each has advantages, but all are slow because they entail inspection of entire images. Furthermore none of these methods have been able to define camera break and gradual transition, which are basic concepts for partitioning. In this paper, we attempt to define camera break. Then, based on our definition and probability analysis, we propose a new video partitioning algorithm, called NET Comparison (NC), which compares the pixels along predefined net lines. In this way, only part of the image is inspected during classification. We compare the effectiveness of our method with other algorithms such as pairwise, likelihood and histogram comparisons, evaluating them on the basis of a large set of varied image sequences that include camera movements, zooming, moving objects, deformed objects and video with degraded image quality. Both gray-level and HSV images were tested and our method out-performed existing approaches in speed and accuracy. On average, our method processes images two to three times faster than the best existing approach.© (1995) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.
TL;DR: In this paper, the authors explore the possibility and flexibility of selecting a subset of all possible pairs of comparisons to assess priority weights of the decision criteria, and provide a framework for unifying existing assessment methods, and reveal the intricate relationships among the methods over the subsets.
01 Jan 1995
TL;DR: In this paper, the authors developed a method of replication coupled with the use of the quality confidence intervals in order to generate invigorating debates on a particular issue before weight assignments are made.
01 Nov 1995
TL;DR: In this article, the use of the analytic hierarchy process (AHP) to assess the contributions of engineering student team members is described, where qualitative factors are of prime of importance.
Abstract: Employment of the analytic hierarchy process (AHP) to assess the contributions of engineering student team members is described. The students perform this assessment as well as the instructor. The assessments are correlated with the individual team role assignments to making a final determination of the contribution of the individual to the team effort. Using this evaluation technique appears to provide candid student peer input for evaluations. AHP can be characterized as a multi-criteria decision technique in which qualitative factors are of prime of importance. A model of the problem is developed using a hierarchical representation. At the top of the hierarchy is the overall goal or prime objective one is seeking to fulfil. The succeeding lower levels then represent the progressive decomposition of the problem. The knowledgeable parties complete a pair-wise comparison of all entries in each level relative to each of the entries in the next higher level of the hierarchy. The composition of these judgments fixes the relative priority of the entities at the lowest level relative to achieving the top-most objective. A description of AHP for teams within a production engineering class is described. First, the lack of success with traditional student questionnaires to assess team performance is described, followed by a description of what appear to be more meaningful results when AHP is used. Finally, several complicating factors associated with this experiment, some tentative conclusions and a recommendation for continued investigation of the use of AHP for student evaluations are described.
TL;DR: In this paper, the authors compare three different comparison procedures and compare them to two methods for comparing means, i.e., the trimmed mean and the sample mean, which is a measure of location having a standard error that is relatively unaffected by heavy tails and outliers.
Abstract: Two common goals when choosing a method for performing all pairwise comparisons of J independent groups are controlling experiment wise Type I error and maximizing power. Typically groups are compared in terms of their means, but it has been known for over 30 years that the power of these methods becomes highly unsatisfactory under slight departures from normality toward heavy-tailed distributions. An approach to this problem, well-known in the statistical literature, is to replace the sample mean with a measure of location having a standard error that is relatively unaffected by heavy tails and outliers. One possibility is to use the trimmed mean. This paper describes three such multiple comparison procedures and compares them to two methods for comparing means.
TL;DR: The authors analyzes the use of a version of lexical maximim strategies, called protective behavior, in two-sided matching models and shows that truth-telling is the unique form of protective behavior in two such mechanisms.
Posted Content•
TL;DR: In this paper, a simple approach and a linear programming approach are formulated and analyzed in solving the problem of combining the comparisons of the two matrices in order to derive the priorities of the elements considered in both matrices.
Abstract: One of the most difficult issues in many real-life decisionmaking problems is how to estimate the pertinent data. An approach which uses pairwise comparisons was proposed by Saaty and is widely accepted as an effective way of determining these data. Suppose that two matrices with pairwise comparisons are available. Furthermore, suppose that there is an overlapping of the elements compared in these two matrices. The problem examined in this paper is how to combine the comparisons of the two matrices in order to derive the priorities of the elements considered in both matrices. A simple approach and a linear programming approach are formulated and analyzed in solving this problem. Computational results suggest that the LP approach, under certain conditions, is an effective way for dealing with this problem. The proposed approach is of critical importance because it can also result in a reduction of the total required number of comparisons.
TL;DR: In this paper, an operational method for deriving a linear ranking of alternatives from repeated paired comparisons of the alternatives is presented, where intransitivities in the observed preferences are cleared away by the introduction of decision errors of varying importance.
TL;DR: A cardinal choice model is presented for analysing the preferences of fishermen and for comparing fishing sites at the Ruunaa sports and recreation fishing area in eastern Finland to help fishermen who have little or no prior information about the area to choose the best possible fishing site in line with their preferences.
TL;DR: An improved version of the original algorithm is introduced that avoids costly dot-matrix multiplications and runs in O(n^3 . L^3) time (L is the length of the longest sequence and n is the number of sequences).
10 Jul 1995
TL;DR: The iterative pairwise mutually best merge (IPMBM) segmentation algorithm extracts image regions characterized by low within-segment variance, and employs a new metric, called normalized vector distance (NVD), to perform a normalized comparison between a pair of multivalued vectors.
Abstract: The iterative pairwise mutually best merge (IPMBM) segmentation algorithm extracts image regions characterized by low within-segment variance, IPMBM employs a new metric, called normalized vector distance (NVD), to perform a normalized comparison between a pair of multivalued vectors. IPMBM is robust and easy to use since only two parameters, both having an intuitive physical meaning, must be user-defined. Experimental results demonstrate that IPMBM is effective in real applications.
TL;DR: This paper elucidates the main problems associated with estimation in the polytomous logistic regression model and introduces a modified procedure in which all pairwise binary logistic models are fitted and combined to construct a single vector of estimates.
TL;DR: The analytic hierarchy process (AHP) of T.L. L. Saaty as mentioned in this paper has been criticised for inconsistency on the part of the decision maker in the case of pairwise comparison.
Abstract: Prioritization strategies characterized by pairwise comparison provide some inconsistency on the part of the decision maker. Such a tolerance is desirable and even necessary given the imprecise nature of human decision-making processes. Methods based on the analytic hierarchy process (AHP) of T. L. Saaty have to date used a somewhat arbitrary measure of consistency. A consideration of the distributions of randomly generated judgment matrices highlights the need for greater refinement. This paper illustrates the problem as manifested in both the AHP and the modified AHP and points the way to a solution in harmony with acknowledged statistical procedures
01 Jan 1995
TL;DR: The class of pairwise difference priors is bound to have a substantial impact on future applied work, and the construction of simultaneous credible regions based on MCMC output is also valuable.
Abstract: We congratulate the authors on a magnificent paper, providing a nicely paced introduction to Markov chain Monte Carlo and its applications, together with several new ideas. In particular the class of pairwise difference priors is bound to have a substantial impact on future applied work. Other ideas given less prominence in the paper are also valuable, for example, the construction of simultaneous credible regions based on MCMC output. There are several issues which we wish to comment on in detail.
TL;DR: This paper uses two medical examples to motivate a discussion illustrating that, when the marginal model contains only covariates specific to many observations within a cluster, the estimating equations used for the marginal analyses can be formulated in terms of proportions rather than binary indicator variables, while still modelling the pairwise associations with covariate specific to individual measurements within the cluster.
Abstract: Several researchers have proposed generalized estimating equation (GEE) approaches for the analysis of clustered binary or multicategory response data. When large numbers of observations are taken within clusters, these GEE methods can require inversion of extremely large covariance matrices. However, in many applications, only covariates specific to large groups of observations within the cluster are of interest for the marginal model, whereas covariates specific to individual measurements within the cluster are important in modelling the pairwise associations. In this paper, we use two medical examples to motivate a discussion illustrating that, when the marginal model contains only covariates specific to many observations within a cluster, the estimating equations used for the marginal analyses can be formulated in terms of proportions rather than binary indicator variables, while still modelling the pairwise associations with covariates specific to individual measurements within the cluster. This approach can reduce the computational burden that is inherent in the analysis of large clusters, while still allowing the potential gains in efficiency for the marginal analysis that can be obtained by modelling the pairwise associations
Journal Article•
TL;DR: Parallel implementations described in this work follow different message-passing strategies and capitalise on the fact that the vector of weights for each matrix can be calculated independently.