Consensus Ranking Research Articles

Faculty consensus on competitiveness for the new competency-based emergency medicine standardized letter of evaluation.

Emergency medicine (EM) has introduced a new, competency-based standardized letter of evaluation (SLOE) template. While a previous version of the SLOE has been shown to promote a high degree of faculty consensus regarding competitiveness, this has not been shown for the new SLOE template. The objective was to evaluate faculty consensus on competitiveness for the new EM SLOE 2.0. Fifty mock SLOE 2.0 letters using the new template were drafted and sent to a group of experienced EM educators. The 50 letters were ranked by the experienced faculty as well as a point-based prediction model and a regression model and the results were compared. Faculty consensus on competitiveness remained strong when using the new SLOE 2.0 format. The points-based prediction model and regression model both demonstrated a high level of agreement with faculty consensus rankings for the SLOE 2.0. Introduction of the new, competency-based SLOE 2.0 format did not have a deleterious effect on faculty consensus rankings of competitiveness.

Quantitative Analysis of Aeolian Sand Provenance: A Comprehensive Analysis in the Otindag Dune Field, Central Inner Mongolia, China

The identification and quantification of aeolian sand contributions are essential for understanding the formation of dune fields and mechanisms of modern surface processes. In the present study, we take aeolian sand in the Otindag dune field (hereafter, often referred to as, simply, Otindag) as the research object. The dune field’s immediate source is quantitatively identified based on heavy minerals and the Conservativeness Index (CI), Consensus Ranking (CR), and the Consistent Tracer Selection (CTS) method. The primary source area of the aeolian sand was found to be from the northwestern, upwind area of the Otindag (59 ± 14%), followed by the Yinshan Mountain (17 ± 10%) and the lake basin (23 ± 12%). The proposed sediment transport model elucidates that sediments from the upwind of the Otindag are directly transported from the northwest to the Otindag, where they are deposited. Materials from the southern Yinshan Mountains are carried by rivers to the southern edge of the Otindag, where they are subsequently transported by wind and ultimately deposited. The lake deposits within the Otindag also contribute to the aeolian sand supply under the influence of wind. This study demonstrates that the fingerprinting techniques of CI, CR, and CTS serve as successful strategies for conducting quantitative provenance research in dune fields.

Ethnobotanical study of medicinal plants of the Abohoy Gara Mountains, Northern Ethiopia

IntroductionFor many years, medicinal plants have been used in Ethiopia to manage a wide range of diseases. Documenting use of the plants for traditional medicine is necessary bases for obtaining new lead bioactive plant based natural products for the discovery of allopathic drugs, preserving the associated local knowledge and medicinal plants. This study aimed to document ethnobotanical knowledge associated with medicinal plants, parts used, route of preparations, and mode of administration. MethodsThe study was conducted from February 2017 to March 2018. The data were collected by interviewing 177 (139 male and 38 female) informants. Semistructured interviews, group discussions, guided field waks and quantitative approaches, such as informant consensus factors and ranking were used to collect and quantify the data. The data were analysed using Microsoft Excel, t test and analysis of variance (ANOVA). ResultsA total of 57 medicinal plant species belonging to 51 genera in 36 families were used by local people to prevent various ailments. The Asteraceae was the dominant family contributing the greatest number of medicinal plants. Leaves were the most widely used plant parts for treating human ailments (20, 56.7 %). Most of the collected traditional medicinal plants were used for treating human ailments directly (11, 28 %), while crushing was the most common method of preparation for treating livestock ailments (42, 70.2 %). Half of the herbal medicines were prepared for internal use, accounting for 31 (54.4 %). The findings showed that Cynoglossum coeruleum was the most preferred species for treating febrilosis, accounting for 33 cases. The highest informant consensus factor (0.83) was obtained for diseases related to febrilosis, and the lowest (0.56) was obtained for diseases associated with livestock (leech infestation and ectoparasites). ConclusionsThis study showed that medicinal plants in the study area, coupled with rich indigenous knowledge, were used by a large member of the population and are the most important means of treating some common ailments. However, medicinal plants in the study district are under pressure due to both natural and anthropogenic influences. Hence, the declining medicinal plants in the area require the application of complementary on-site and off-site conservation approaches.

RAbDesFlow: a novel workflow for computational recombinant antibody design for healthcare engineering.

Recombinant antibodies (rAbs) have emerged as a promising solution to tackle antigen specificity, enhancement of immunogenic potential and versatile functionalization to treat human diseases. The development of single chain variable fragments has helped accelerate treatment in cancers and viral infections, due to their favorable pharmacokinetics and human compatibility. However, designing rAbs is traditionally viewed as a genetic engineering problem, with phage display and cell free systems playing a major role in sequence selection for gene synthesis. The process of antibody engineering involves complex and time-consuming laboratory techniques, which demand substantial resources and expertise. The success rate of obtaining desired antibody candidates through experimental approaches can be modest, necessitating iterative cycles of selection and optimization. With ongoing advancements in technology, in silico design of diverse antibody libraries, screening and identification of potential candidates for in vitro validation can be accelerated. To meet this need, we have developed rAbDesFlow, a unified computational workflow for recombinant antibody engineering with open-source programs and tools for ease of implementation. The workflow encompasses five computational modules to perform antigen selection, antibody library generation, antigen and antibody structure modeling, antigen-antibody interaction modeling, structure analysis, and consensus ranking of potential antibody sequences for synthesis and experimental validation. The proposed workflow has been demonstrated through design of rAbs for the ovarian cancer antigen Mucin-16 (CA-125). This approach can serve as a blueprint for designing similar engineered molecules targeting other biomarkers, allowing for a simplified adaptation to different cancer types or disease-specific antigens.

Identification of core competencies for exercise oncology professionals: A Delphi study of United States and Australian participants.

Integration of exercise into standard oncology care requires a highly skilled workforce of exercise professionals; however, competency requirements have not kept pace with advancements in the field. Therefore, the aim of this study was to obtain consensus on core competencies required for an exercise professional to be qualified to work with adults undergoing active cancer treatment. A three-round modified electronic Delphi process was used. In Round 1, an international group of 64 exercise oncology stakeholders (i.e., exercise oncology professionals (n = 29), clinical referrers (n = 21), and people with lived experience (n = 14)) responded to open-ended prompts eliciting perspectives regarding competencies needed for an exercise oncology professional to work with adults receiving active cancer treatment. Subsequently, only exercise oncology professionals participated, ranking the importance of competencies. In Round 2, professionals received summary feedback, ranked new competencies generated from open-ended responses, and reranked competencies not reaching consensus. In the final round, professionals finalized consensus ranking and rated frequency and mastery level for each. Consensus was reached on 103 core competencies required for exercise professionals to be qualified to deliver care to adults undergoing active cancer treatment. The core competencies represent 10 content areas and reflect the needs of clinical referrers and people with lived experience of receiving cancer treatment. The core competencies identified reflect significant advancements in the field of exercise oncology. Results will underpin the development of education, certification, and employment requirements for exercise oncology professionals, providing a critical step toward achieving routine integration of exercise into standard oncology care.

Measuring and Predicting Faculty Consensus Rankings of Standardized Letters of Evaluation.

Background Standardized letters of evaluation (SLOE) are becoming more widely incorporated into the residency application process to make the letter of recommendation, an already critical component in a residency application packet, more objective. However, it is not currently known if the reviewers of these letters share consensus regarding the strength of an applicant determined by their SLOE. Objective We measured the level of faculty agreement regarding applicant competitiveness as determined by SLOEs and the ability of 2 algorithms to predict faculty consensus rankings. Methods Using data from the 2021-2022 Match cycle from the Council of Residency Directors in Emergency Medicine SLOE Database as a blueprint, authors created 50 fictional SLOEs representative of the national data. Seven faculty then rated these SLOEs in order of applicant competitiveness, defined as suggested rank position. Consensus was evaluated using cutoffs established a priori, and 2 prediction models, a point-based system and a linear regression model, were tested to determine their ability to predict consensus rankings. Results There was strong faculty consensus regarding the interpretation of SLOEs. Within narrow windows of agreement, faculty demonstrated similar ranking patterns with 83% and 93% agreement for "close" and "loose" agreement, respectively. Predictive models yielded a strong correlation with the consensus ranking (point-based system r=0.97, linear regression r=0.97). Conclusions Faculty displayed strong consensus regarding the competitiveness of applicants via SLOEs, adding further support to the use of SLOEs for selection and advising. Two models predicted consensus competitiveness rankings with a high degree of accuracy.

Effective signal reconstruction from multiple ranked lists via convex optimization

The ranking of objects is widely used to rate their relative quality or relevance across multiple assessments. Beyond classical rank aggregation, it is of interest to estimate the usually unobservable latent signals that inform a consensus ranking. Under the only assumption of independent assessments, which can be incomplete, we introduce indirect inference via convex optimization in combination with computationally efficient Poisson Bootstrap. Two different objective functions are suggested, one linear and the other quadratic. The mathematical formulation of the signal estimation problem is based on pairwise comparisons of all objects with respect to their rank positions. Sets of constraints represent the order relations. The transitivity property of rank scales allows us to reduce substantially the number of constraints associated with the full set of object comparisons. The key idea is to globally reduce the errors induced by the rankers until optimal latent signals can be obtained. Its main advantage is low computational costs, even when handling n<<p\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$n < < p$$\\end{document} data problems. Exploratory tools can be developed based on the bootstrap signal estimates and standard errors. Simulation evidence, a comparison with the state-of-the-art rank centrality method, and two applications, one in higher education evaluation and the other in molecular cancer research, are presented.

Heuristic Search for Rank Aggregation with Application to Label Ranking

Rank aggregation combines the preference rankings of multiple alternatives from different voters into a single consensus ranking, providing a useful model for a variety of practical applications but posing a computationally challenging problem. In this paper, we provide an effective hybrid evolutionary ranking algorithm to solve the rank aggregation problem with both complete and partial rankings. The algorithm features a semantic crossover based on concordant pairs and an enhanced late acceptance local search method reinforced by a relaxed acceptance and replacement strategy and a fast incremental evaluation mechanism. Experiments are conducted to assess the algorithm, indicating a highly competitive performance on both synthetic and real-world benchmark instances compared with state-of-the-art algorithms. To demonstrate its practical usefulness, the algorithm is applied to label ranking, a well-established machine learning task. We additionally analyze several key algorithmic components to gain insight into their operation. History: Accepted by Erwin Pesch, Area Editor for Heuristic Search &amp; Approximation Algorithms. Funding: This work was supported by the National Natural Science Foundation of China [Grant 72371157] and Shanghai Pujiang Programme [Grant 23PJC069]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2022.0019 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2022.0019 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .

Structural connectivity from DTI to predict mild cognitive impairment in de novo Parkinson’s disease

BackgroundEarly detection of Parkinson's disease (PD) patients at high risk for mild cognitive impairment (MCI) can help with timely intervention. White matter structural connectivity is considered an early and sensitive indicator of neurodegenerative disease. ObjectivesTo investigate whether baseline white matter structural connectivity features from diffusion tensor imaging (DTI) of de novo PD patients can help predict PD-MCI conversion at an individual level using machine learning methods. MethodsWe included 90 de novo PD patients who underwent DTI and 3D T1-weighted imaging. Elastic net-based feature consensus ranking (ENFCR) was used with 1000 random training sets to select clinical and structural connectivity features. Linear discrimination analysis (LDA), support vector machine (SVM), K-nearest neighbor (KNN) and naïve Bayes (NB) classifiers were trained based on features selected more than 500 times. The area under the ROC curve (AUC), accuracy (ACC), sensitivity (SEN) and specificity (SPE) were used to evaluate model performance. ResultsA total of 57 PD patients were classified as PD-MCI nonconverters, and 33 PD patients were classified as PD-MCI converters. The models trained with clinical data showed moderate performance (AUC range: 0.62–0.68; ACC range: 0.63–0.77; SEN range: 0.45–0.66; SPE range: 0.64–0.84). Models trained with structural connectivity (AUC range, 0.81–0.84; ACC range, 0.75–0.86; SEN range, 0.77–0.91; SPE range, 0.71–0.88) performed similar to models that were trained with both clinical and structural connectivity data (AUC range, 0.81–0.85; ACC range, 0.74–0.85; SEN range, 0.79–0.91; SPE range, 0.70–0.89). ConclusionsBaseline white matter structural connectivity from DTI is helpful in predicting future MCI conversion in de novo PD patients.

Integration of circulating microRNAs and transcriptome signatures identifies early-pregnancy biomarkers of preeclampsia.

MicroRNAs (miRNAs) have been implicated in the pathobiology of preeclampsia, a common hypertensive disorder of pregnancy. In a nested matched case-control cohort within the Vitamin D Antenatal Asthma Reduction Trial (VDAART), we previously identified peripheral blood mRNA signatures related to preeclampsia and vitamin D status (≤30ng/mL) during gestation from 10 to 18weeks, using differential expression analysis. Using quantitative PCR arrays, we conducted profiling of circulating miRNAs at 10-18weeks of gestation in the same VDAART cohort to identify differentially expressed (DE) miRNAs associated with preeclampsia and vitamin D status. For the validation of the expression of circulating miRNA signatures in the placenta, the HTR-8/SVneo trophoblast cell line was used. Targets of circulating miRNA signatures in the preeclampsia mRNA signatures were identified by consensus ranking of miRNA-target prediction scores from four sources. The connected component of target signatures was identified by mapping to the protein-protein interaction (PPI) network and hub targets were determined. As experimental validation, we examined the gene and protein expression of IGF1R, one of the key hub genes, as a target of the DE miRNA, miR-182-5p, in response to a miR-182-5p mimic in HTR-8/SVneo cells. Pregnant women with preeclampsia had 16 circulating DE miRNAs relative to normal pregnancy controls that were also DE under vitamin D insufficiency (9/16=56% upregulated, FDR<.05). Thirteen miRNAs (13/16=81.3%) were detected in HTR-8/SVneo cells. Overall, 16 DE miRNAs had 122 targets, of which 87 were unique. Network analysis demonstrated that the 32 targets of DE miRNA signatures created a connected subnetwork in the preeclampsia module with CXCL8, CXCL10, CD274, MMP9 and IGF1R having the highest connectivity and centrality degree. In an in vitro validation experiment, the introduction of an hsa-miR-182-5p mimic resulted in significant reduction of its target IGF1R gene and protein expression within HTR-8/SVneo cells. The integration of the circulating DE miRNA and mRNA signatures associated preeclampsia added additional insights into the subclinical molecular signature of preeclampsia. Our systems and network biology approach revealed several biological pathways, including IGF-1, that may play a role in the early pathophysiology of preeclampsia. These pathways and signatures also denote potential biomarkers for the early stages of preeclampsia and suggest possible preventive measures.

Inference on Consensus Ranking of Distributions

Instead of testing for unanimous agreement, I propose learning how broad of a consensus favors one distribution over another (of earnings, productivity, asset returns, test scores, etc.). Specifically, given a sample from each of two distributions, I propose statistical inference methods to learn about the set of utility functions for which the first distribution has higher expected utility than the second distribution. With high probability, an “inner” confidence set is contained within this true set, while an “outer” confidence set contains the true set. Such confidence sets can be formed by inverting a proposed multiple testing procedure that controls the familywise error rate. Theoretical justification comes from empirical process results, given that very large classes of utility functions are generally Donsker (subject to finite moments). The theory additionally justifies a uniform (over utility functions) confidence band of expected utility differences, as well as tests with a utility-based “restricted stochastic dominance” as either the null or alternative hypothesis. Simulated and empirical examples illustrate the methodology.

A unifying rank aggregation framework to suitably and efficiently aggregate any kind of rankings

The aggregation of multiple rankings into a consensus ranking is a crucial task in various domains such as search engine results or user-based ratings. This task poses significant challenges due to its inherent complexity. The complexity of the problem stems not only from the need for exactness and efficiency, but also from the diversity of real-world scenarios, which often involve incomplete rankings and ties.Most existing methods propose a specific way to aggregate rankings. However, these methods often do not take into account different real use-case scenarios which can impact the relevance of their final result, as the congruence between the aggregated output and the expected outcome inherently depends on the context. To address the issue of context-dependency in ranking aggregation, we introduce a unifying framework that subsumes a variety of generalizations of the Kemeny score for incomplete rankings with ties and enables the design of new ones if a specific context requires it. Our framework is parameterized, allowing for different behaviors depending on the specific use case.We provide a broader scope of application to the methods encompassed by our approach, augmenting them with a larger theoretical and algorithmic structure. We establish an axiomatic study to better understand each method within our framework and present an algorithmic approach that includes exact methods, partitioning algorithms, and heuristics.Finally, we demonstrate the practical relevance of our approach through an empirical study on both real and synthetic datasets. Notably, the synthetic datasets are generated based on devised real-world scenarios, highlighting the context-dependent applicability of different Kemeny-based rank aggregation methods within our framework.

ChemFlow_py: a flexible toolkit for docking and rescoring.

The design of accurate virtual screening tools is an open challenge in drug discovery. Several structure-based methods have been developed at different levels of approximation. Among them, molecular docking is an established technique with high efficiency, but typically low accuracy. Moreover, docking performances are known to be target-dependent, which makes the choice of the docking program and corresponding scoring function critical when approaching a new protein target. To compare the performances of different docking protocols, we developed ChemFlow_py, an automated tool to perform docking and rescoring. Using four protein systems extracted from DUD-E with 100 known active compounds and 3000 decoys per target, we compared the performances of several rescoring strategies including consensus scoring. We found that the average docking results can be improved by consensus ranking, which emphasizes the relevance of consensus scoring when little or no chemical information is available for a given target. ChemFlow_py is a free toolkit to optimize the performances of virtual high-throughput screening (vHTS). The software is publicly available at https://github.com/IFMlab/ChemFlow_py .

Informative Priors for the Consensus Ranking in the Bayesian Mallows Model

The aim of this work is to study the problem of prior elicitation for the consensus ranking in the Mallows model with Spearman’s distance, a popular distance-based model for rankings or permutation data. Previous Bayesian inference for such a model has been limited to the use of the uniform prior over the space of permutations. We present a novel strategy to elicit informative prior beliefs on the location parameter of the model, discussing the interpretation of hyper-parameters and the implication of prior choices for the posterior analysis.

Static and dynamic source identification of trace elements in river and soil environments under anthropogenic activities in the Haraz plain, Northern Iran

Unsustainable human activities have disrupted the natural cycle of trace elements, causing the accumulation of chemical pollutants and making it challenging to determine their sources due to interwoven natural and human-induced processes. A novel approach was introduced for identifying the sources and for quantifying the contribution of trace elements discharge from rivers to soils. We integrated fingerprinting techniques, soil and sediment geochemical data, geographically weighted regression model (GWR) and soil quality indices. The FingerPro package and the state-of-the-art tracer selection techniques including the conservative index (CI) and consensus ranking (CR) were used to quantify the relative contribution of different upland sub-watersheds in trace element discharge soil. Our analysis revealed that off-site sources (upland watersheds) and in-site sources (land use) both play an important role in transferring trace elements to the Haraz plain (northern Iran). The unmixing model's results suggest that the Haraz sub-watersheds exhibit a higher contribution to trace elements transfer in the Haraz plain, and therefore, require greater attention in terms of implementing soil and water conservation strategies. However, it is noteworthy that the Babolroud (adjacent to Haraz) exhibited a better performance of the model. A spatial correlation between certain heavy metals, such as As and Cu, and rice cultivation existed. Additionally, we found a significant spatial correlation between Pb and residential areas, particularly in the Amol region. Our result highlights the importance of using advanced spatial statistical techniques, such as GWR, to identify subtle but critical associations between environmental variables and sources of pollution. The methodology used comprehensively identifies dynamic trace element sourcing at the watershed scale, allowing for pollutant source identification and practical strategies for soil and water quality control. Tracer selection techniques (CI and CR) based on conservatives and consensus improve unmixing model accuracy and flexibility for precise fingerprinting.

Measuring and predicting faculty consensus rankings of Standardized Letters of Evaluation

Measuring and predicting faculty consensus rankings of Standardized Letters of Evaluation

Ethnobotanical Study on Wild Edible Plants in Metema District, Amhara Regional State, Ethiopia

Wild edible plants are vital for the survival and sustainable livelihoods of rural people of Ethiopia. Thus, this study compiled wild edible plants, their use, threats, and management practices in Metema District, northwestern Ethiopia. Eight sample kebeles were selected purposively based on vegetation coverage and key informants availability. A total of 128 informants were selected using purposive and random sampling techniques. Data were collected using individual interviews, guided field walks, focus group discussion, and market surveys through semistructured questionnaires. Ethnobotanical data collected from the informants were summarized by descriptive statistics, and further verified by using informant consensus, preference, and direct matrix ranking. Knowledge difference among age and sex groups was evaluated by independent sample t-test. A total of 44 wild edible plant species distributed in 34 genera and 25 families were documented. Most (88.64%) of these plant species were found in the wild habitat. Fabaceae and Moraceae accounted for higher proportion of edible plant species. Trees were the dominant habit (59.09%). From the total recorded wild edible plants, 33 (75%) species were used as supplementary foods and 11 (25%) species were used during famine. Fruits were the most edible plant parts (66%) and raw fresh forms were the main conditions of consumption (81.8%). Diospyros abyssinica was the most cited (60.94%) and first ranked. Some edible species such as Adansonia digitata and Balanites aegyptiaca were marketable. Ziziphus spina-christi was found the most multipurpose wild edible plant species. Most of the species (33, 75%) were used as animal fodder followed with traditional medicines (25, 56.82%) and firewood (20, 45.45%). Tamarindus indica, Moringa stenopetala, Balanites aegyptiaca, Grewia ferruginea, Corchorus olitorius, and Cordia africana had nutraceutical values. Significant knowledge differences ( P &lt; 0.05 ) were obtained among sexes and age groups of informants on the number of wild edible plant species they listed. As a result of their multiple roles, wild edible plants are threatened by various anthropogenic activities. Despite this, Metema District still supports good numbers of wild edible plants from which the poor inhabitants complement their basic needs by consuming and marketing them. For sustainable utilization, conservation, value addition, and market linkage practices shall be strengthened to improve the livelihoods of local people and sustainable forest management.

Sediment source tracing during flood events in the Huangfu River basin in the northern Loess Plateau, China

Identifying critical soil erosion areas and sediment sources are essential for sediment control and water resources management in the watershed, especially in the severely eroded northern Loess Plateau, China. In this study, we collected suspended sediment samples at the hydrological stations to identify the sediment source dynamics during flood events in a highly erodible basin, the Huangfu River basin in the Northern Loess Plateau. A total of 6 flood events were sampled during 2016 and 2018 for investigation, the flow discharge was featured by a rapid rising limb and a slow falling limb. Compared with the sediment sources, the silt fraction in the suspended sediments was enriched and the sand fraction in the suspended sediment was depleted. The dominant particle grain size of suspended sediments was coarse sand (56.89%), higher than silt (41.81%) and clay (1.3%). The sediment sources were estimated using the FingerPro unmixing model based on the conservativeness index (CI) and consensus ranking (CR) methods. The results of the unmixing model reveal that the bare land was the predominant source of suspended sediments, ranging from 40% to 57%, which was much higher than farmland (27 ∼ 31%) and grassland (16 ∼ 25%). Our findings imply that the bare land (weathered sandstone area) is the critical region for soil loss control in the Huangfu River basin. A combination of vegetation restoration measures and engineering measures are proposed to improve the sustainable development of ecological environment in the basin.

Group decision-making method of commercial bank risk monitoring based on PCbHA iteration

The steady development of commercial banks plays a key role in maintaining the healthy development of the economy. The ability to judge financial risks is a reflection of the comprehensive risk management level of a commercial bank, and it is also an important criterion for measuring its competitiveness and operational stability. Based on the analysis of economic development laws and reference to relevant literature, this paper screened out the eight most representative risk evaluation measurement indicators of commercial banks, ranked these indicators in preference according to expert opinions, established a group decision-making model, and then obtained the consensus ranking by using the least divergence method. The PCbHA method was used to check the consistency of the results, modify the expert opinions, iteratively calculate, and finally construct the importance ranking of commercial bank risk indicators. This paper discusses the construction of an evaluation system based on the perspective of risk management to enrich and improve the risk management content of commercial banks, enhance the risk prevention and control ability, and provide suggestions for the prevention and management of risks in commercial banks.

Analytic hierarchy process and data envelopment analysis: A match made in heaven

Analytic hierarchy process (AHP) and data envelopment analysis (DEA) are two popular decision science methods with many business, science, and engineering applications. AHP is a Multi-Attribute Decision-Making (MADM) method for prioritizing alternatives, and DEA is a non-parametric method for estimating production frontiers. Each method has known strengths and weaknesses, and the strengths in one method can overcome the weaknesses in the other. We study several strategies to diminish the weaknesses of DEA with the strengths of pairwise comparisons in AHP. We tackle the low discrimination power inherent to conventional DEA methods. AHP, with its pairwise comparison capability, has been consistently used to increase the discrimination power and accuracy in DEA. We propose and evaluate several new hybrid MADM-DEA models of different computational complexity and consistency, including combinations of the best-worst method (BWM) and its variants with DEA as well as a novel method composed of Measuring Attractiveness by a Categorical Based Evaluation Technique (MACBETH) and DEA. We further develop a new technique for evaluating the similarity among multiple ranking results in MADM. The new simple but powerful technique is called Rank Absolute Deviation (RAD) and is inspired by the mean absolute deviation method. Several numerical examples and a real-world problem are used to demonstrate the applicability and efficacy of the new BWM-DEA, MACBETH-DEA, and RAD methods proposed in this study. We illustrate how less computationally demanding MADM-DEA techniques provide rankings that are highly correlated with the benchmark DEA-AHP and different consensus ranking models.