Ranking Order Research Articles

Feature importance explanation of prosthetic valve endocarditis through machine learning via SHAP

Abstract Introduction Diagnosing prosthetic valve endocarditis poses a significant challenge in daily clinical practice. Accurately identifying these patients is vital due to its treatment implications. Clinical and imaging assessment of suspected cases yields a substantial number of interrelated variables, which can be analyzed through machine learning techniques in order to improve diagnostic accuracy. Feature ranking facilitates identification and interpretability of relevant clinical and imaging variables in this task. Purpose To utilize machine learning feature ranking and modeling in order to improve interpretable identification of patients with prosthetic valve endocarditis. Methods 89 cases defined by the endocarditis team and 173 controls were analyzed. We implemented the SHapley Additive exPlanations (SHAP) method to evaluate feature importance. 5-fold cross-validated Xgboost models were trained and independently tested with additive clinical, echocardiographic and 18F-FDG PET/CT data. Performance was evaluated through confusion matrices, F1-scores and AUCs. These were compared against conventional logistic regression of the Duke Criteria classification. Results Mean age was 59.6 ± 15.7, with 29.7% women. 14 clinical (demographic, laboratory and cultures), 2 echocardiographic, and 5 PET/CT features were included. The XGBoost model trained with all additive features demonstrated an AUC of 0.91±0.1 and outperformed simpler models as well as the Duke classification model (AUC=0.85). The clinical variables alone achieved an AUC of 0.85±0.11, while integration with echocardiography enhanced its performance to 0.90±0.04 (Fig. 1). Feature ranking demonstrated that antibiotic treatment duration, positive blood cultures, valve type, weight and CRP were the most relevant clinical variables, while the presence of vegetation or abscess, and abnormal uptake and uptake focality on PET/CT represented the most relevant imaging variables (Fig. 2). Conclusion Machine learning-based feature ranking and modeling may significantly enhance identification of patients with prosthetic valve endocarditis beyond standard clinical criteria. Feature selection offers interpretable performance in identifying these cases, model testing in clinical decision support may be warranted.Fig. 1:Comparative AUC Analysis.Fig. 2:SHAP feature importance.

Adversarial Attack and Defense in Deep Ranking.

Deep Neural Network classifiers are vulnerable to adversarial attacks, where an imperceptible perturbation could result in misclassification. However, the vulnerability of DNN-based image ranking systems remains under-explored. In this paper, we propose two attacks against deep ranking systems, i.e., Candidate Attack and Query Attack, that can raise or lower the rank of chosen candidates by adversarial perturbations. Specifically, the expected ranking order is first represented as a set of inequalities. Then a triplet-like objective function is designed to obtain the optimal perturbation. Conversely, an anti-collapse triplet defense is proposed to improve the ranking model robustness against all proposed attacks, where the model learns to prevent the adversarial attack from pulling the positive and negative samples close to each other. To comprehensively measure the empirical adversarial robustness of a ranking model with our defense, we propose an empirical robustness score, which involves a set of representative attacks against ranking models. Our adversarial ranking attacks and defenses are evaluated on MNIST, Fashion-MNIST, CUB200-2011, CARS196, and Stanford Online Products datasets. Experimental results demonstrate that our attacks can effectively compromise a typical deep ranking system. Nevertheless, our defense can significantly improve the ranking system's robustness and simultaneously mitigate a wide range of attacks.

Evaluation of Green Suppliers Based on a Hybrid Weighting Approach and GRA Method

To align with the national dual carbon strategy and facilitate low-carbon production for enterprises, this paper incorporates the concept of green supply chain management and develops a scientifically sound evaluation index system for green suppliers by comprehensively considering five factors: quality, availability, cost, environmental management performance, and technology. The analytic hierarchy process and entropy weight method are employed to assign subjective and objective weights to each evaluation index while the grey correlation degree analysis method is utilized to rank and sort suppliers in order to select the optimal green supplier. Finally, an empirical example is presented to verify the reliability of our model.

The Influences of Ties in Preference on Market Matching

The ties of preference and priorities are commonly seen and widely used in daily life. When the agent considers the priority order of some selected objects to be the same, the final matching result may be affected. People sometimes require strict preference ranking in order to facilitate the use of matching models, which can cause controversy and dissatisfaction with people in the matching. In order to effectively solve the above problems, this paper surveys what these impacts are and how they affect the final result of matching. According to the analysis, it is proven that when there is a preference of indifference, Pareto efficiency is unstable and the deferred acceptance algorithm cannot be used for market matching. In addition, the Pareto-improving draft mechanism has proven to be powerful without imposing a strict preference. This mechanism can also avoid the complex process and other negative consequences of handing ties arbitrarily.

Evaluation of factors impacting University-Industry collaboration by Fuzzy AHP and Fuzzy TOPSIS methods

University-industry collaboration (UIC) is viewed as a means to boost innovation through knowledge sharing. The purpose of this study is to investigate the importance ranking of factors influencing UIC and UIC models in the context of Vietnam. To begin, a systematic literature review (SLR) of previous academic studies on the subject of university-industry collaboration was conducted in order to identify the influential factors in UIC. The order ranking of factors and their relevance weights are then determined using the fuzzy AHP approach. The fuzzy TOPSIS is then used to evaluate three models for UIC, including collaboration in scientific research, technology transfer, and training. The results show that the number of scientists and experts and the awareness of the benefits of cooperation are the most important factors in implementing UIC; meanwhile, the experience of cooperation and training activities and student recruitment activities are evaluated less important than others. In addition, in the context of Vietnam, among UIC models, the technology transfer collaboration should be prioritized. Finally, based on the findings, discussions and policy implications are provided to promote UIC activities.

Quantifying holistic capacity response and healthcare resilience in tackling COVID-19: Assessment of country capacity by MCDM.

The resilience of a country during the COVID-19 pandemic was determined based in whether it was holistically prepared and responsive. This resilience can only be identified through systematic data collection and analysis. Historical evidence-based response indicators have been proven to mitigate pandemics like COVID-19. However, most databases are outdated, requiring updating, derivation, and explicit interpretation to gain insight into the impact of COVID-19. Outdated databases do not show a country's true preparedness and response capacity, therefore, it undermines pandemic threat. This study uses up-to-date evidence-based pandemic indictors to run a cross-country comparative analysis of COVID-19 preparedness, response capacity, and healthcare resilience. PROMETHEE-a multicriteria decision making (MCDM) technique-is used to quantify the strengths (positive) and weaknesses (negative) of each country's COVID-19 responses, with full ranking (net) from best to least responsive. From 22 countries, South Korea obtained the highest net outranking value of 0.1945, indicating that it was the most resilient, while Mexico had the lowest (-0.1428). Although countries were underprepared, there was a robust response to the pandemic, especially in developing countries. This study demonstrates the performance and response capacity of 22 key countries to resist COVID-19, from which other countries can compare their statutory capacity ranking in order to learn/adopt the evidence-based responses of better performing countries to improve their resilience.

Implementation of an Arabic spell checker

This paper outlines the implementation of a spell checker for the Arabic language, leveraging the capabilities of NooJ and its functionality, specifically noojapply. In this paper, we shall proceed to provide clear definitions and comprehensive descriptions of several categories of spelling errors. Next, we will provide a comprehensive introduction to the NooJ platform and its command-line utility, noojapply. In the subsequent section, we shall outline the four main phases of our spell checker prototype. We intend to develop a local grammar in NooJ for the purpose of error detection. Afterwards, a morphological grammar and a local grammar will be created in NooJ with the aim of providing an exhaustive list of possible corrections. Following that, a revised algorithm will be employed to arrange these candidates in descending order of ranking. Subsequently, a web user interface will be developed to visually represent our research efforts. Finally, we will proceed to showcase a series of tests and evaluations conducted on our prototype, Al Mudaqiq.

Coarse-to-Fine Contrastive Learning on Graphs.

Inspired by the impressive success of contrastive learning (CL), a variety of graph augmentation strategies have been employed to learn node representations in a self-supervised manner. Existing methods construct the contrastive samples by adding perturbations to the graph structure or node attributes. Although impressive results are achieved, it is rather blind to the wealth of prior information assumed: with the increase of the perturbation degree applied on the original graph: 1) the similarity between the original graph and the generated augmented graph gradually decreases and 2) the discrimination between all nodes within each augmented view gradually increases. In this article, we argue that both such prior information can be incorporated (differently) into the CL paradigm following our general ranking framework. In particular, we first interpret CL as a special case of learning to rank (L2R), which inspires us to leverage the ranking order among positive augmented views. Meanwhile, we introduce a self-ranking paradigm to ensure that the discriminative information among different nodes can be maintained and also be less altered to the perturbations of different degrees. Experiment results on various benchmark datasets verify the effectiveness of our algorithm compared with the supervised and unsupervised models.

Interval-valued intuitionistic fuzzy AROMAN method and its application in sustainable wastewater treatment technology selection

In the theory of interval-valued intuitionistic fuzzy set (IVIFS), the rating/grade of an element is the subset of the closed interval [0, 1], therefore the IVIFS doctrine is more useful for the decision expert to present their judgments in terms of intervals rather than the crisp values. The present work develops an integrated decision-making methodology for evaluating sustainable wastewater treatment technologies within the context of IVIFS. The proposed decision-making framework is divided into three stages. First, some Yager weighted aggregation operators and their axioms are developed to combine the interval-valued intuitionistic fuzzy information. These operators can offer us a flexible way to solve the realistic multi-criteria decision-making problems under IVIFS context. Second, an extension of Symmetry Point of Criterion model is introduced to determine the criteria weights under IVIFS environment. Third, an integrated alternative ranking order model accounting for two-step normalization (AROMAN) approach is proposed from IVIF information perspective. Next, the practicability and efficacy of the developed model is proven by implementing it on a case study of sustainable wastewater treatment technologies evaluation problem with multiple criteria and decision experts. Finally, comparative analysis is discussed to illustrate the consistency and robustness of the obtained outcomes.

Prediction of Binding Pose and Affinity of Nelfinavir, a SARS-CoV-2 Main Protease Repositioned Drug, by Combining Docking, Molecular Dynamics, and Fragment Molecular Orbital Calculations.

A novel in silico drug design procedure is described targeting the Main protease (Mpro) of the SARS-CoV-2 virus. The procedure combines molecular docking, molecular dynamics (MD), and fragment molecular orbital (FMO) calculations. The binding structure and properties of Mpro were predicted for Nelfinavir (NFV), which had been identified as a candidate compound through drug repositioning, targeting Mpro. Several poses of the Mpro and NFV complexes were generated by docking, from which four docking poses were selected by scoring with FMO energy. Then, each pose was subjected to MD simulation, 100 snapshot structures were sampled from each of the generated MD trajectories, and the structures were evaluated by FMO calculations to rank the pose based on binding energy. Several residues were found to be important in ligand recognition, including Glu47, Asp48, Glu166, Asp187, and Gln189, all of which interacted strongly with NFV. Asn142 is presumably regarded to form hydrogen bonds or CH/π interaction with NFV; however, in the present calculation, their interactions were transient. Moreover, the tert-butyl group of NFV had no interaction with Mpro. Identifying such strong and weak interactions provides candidates for maintaining and substituting ligand functional groups and important suggestions for drug discovery using drug repositioning. Besides the interaction between NFV and the amino acid residues of Mpro, the desolvation effect of the binding pocket also affected the ranking order. A similar procedure of drug design was applied to Lopinavir, and the calculated interaction energy and experimental inhibitory activity value trends were consistent. Our approach provides a new guideline for structure-based drug design starting from a candidate compound whose complex crystal structure has not been obtained.

Comparison of Aversion to Visual Dental Stimuli Between Patients and Dentists: A Preliminary Study.

Fearful dental patients often cite various dental instruments or procedures as triggers for their dental fear. Thus, visual dental stimuli provoke anxiety. This preliminary study aimed to assess the level of aversion to visual stimuli in dental patients and compare it with that in dentists. A total of 43 dental patients (25 women, 18 men; average age, 29.9 ± 13.3 years; patient group) and 13 dentists (4 women, 9 men; average age, 28.2 ± 2.0 years; dentist group) were included. All participants had previously undergone dental treatment. The dental fear level was assessed using the self-reported Dental Fear Survey (DFS). Thirty-two images associated with dental treatment were prepared and classified into three categories: dental instruments, dental procedures, and the dental environment. All participants rated their level of disgust toward each image on a visual analog scale with scores ranging from 0 to 100. In the patient group, the disgust ratings for tooth extraction, dental drilling, and local anesthesia were >60, which were significantly different from those in the dentist group (Mann-Whitney U-test, p<0.001, p=0.001, and p=0.001, respectively). The ranking order of the disgust ratings for the 32 images showed significant correlation between the patient and dentist groups (Spearman correlation coefficient, r=0.80, p<0.001). In the patient group, the disgust ratings for dental impressions and the interdental brush, dental light, and dental chair were significantly correlated with DFS scores (r=0.61, p<0.001; r=0.47, p=0.001; r=0.41, p=0.006; and r=0.40, p=0.008, respectively). This study revealed that patients have more negative feelings toward invasive procedures than dentists. However, a significant correlation was identified between the ranking of aversion-provoking dental stimuli by patients and dentists. Furthermore, the level of aversion to several dental-related items that do not cause pain was correlated with the dental fear level.

Cattle tick-borne pathogens sharing the same vectors in Benin and Burkina Faso: variations in prevalence and coinfection patterns

Understanding interactions between pathogens at the level of the individual host and the population in West Africa, may have noteworthy implications for predictions of diseases emergence and disease control programmes. Hence, the current study was aimed at investigating the interactions between Anaplasma marginale, Babesia bigemina and B. bovis in West African cattle. Twelve sentinel steers in each of the eight selected herds were randomly chosen to perform a one-year-long survey designed to monthly collect blood sample and to diagnose possible infections of the steers. This allowed identifying A. marginale as the most prevalent pathogen across the three surveyed regions (range: 0.60-1), followed by B. bigemina (0.24-0.85) and B. bovis (0.10-0.64); the same ranking order of the tick-borne pathogens was observed when considering the apparent duration of cattle infection. Regarding interaction patterns, mainly avoidance was revealed between A. marginale and B. bigemina and between B. bigemina and B. bovis. Such negative interaction seemed to be beneficial to the West African cattle surveyed, helping them to escape from babesiosis caused by B. bovis. Altogether, studied Boophilus tick-borne pathogens negatively interact within cattle in Benin and Burkina Faso. This pattern raises new questions regarding the underlying mechanisms and potential consequences.

Machine Learning Methods as a Cost-Effective Alternative to Physics-Based Binding Free Energy Calculations.

The rank ordering of ligands remains one of the most attractive challenges in drug discovery. While physics-based in silico binding affinity methods dominate the field, they still have problems, which largely revolve around forcefield accuracy and sampling. Recent advances in machine learning have gained traction for protein-ligand binding affinity predictions in early drug discovery programs. In this article, we perform retrospective binding free energy evaluations for 172 compounds from our internal collection spread over four different protein targets and five congeneric ligand series. We compared multiple state-of-the-art free energy methods ranging from physics-based methods with different levels of complexity and conformational sampling to state-of-the-art machine-learning-based methods that were available to us. Overall, we found that physics-based methods behaved particularly well when the ligand perturbations were made in the solvation region, and they did not perform as well when accounting for large conformational changes in protein active sites. On the other end, machine-learning-based methods offer a good cost-effective alternative for binding free energy calculations, but the accuracy of their predictions is highly dependent on the experimental data available for training the model.

New Service Concept and Investor Satisfaction of Aviation Firms in Rivers State

The aim of this study was to ascertain the extent of the relationship between new service concept and investor satisfaction of aviation firms in Rivers State. This study adopted causal research, involving cross sectional survey design and used structured copies of questionnaire. The population of this study comprised of eight (8) airlines registered with the Rivers State ministry of commerce and industry. 20 Copies of questionnaire was distributed to management staff in each of the airlines. The total number of respondents for the study was one hundred and sixty (160). The hypotheses were tested using Spearman Ranking Order Correlation with the aid of SPSS Version 21. The study unveiled a very strong relationship between new service concept and investor satisfaction. This study therefore concluded that increased investor satisfaction can be achieved if aviation firms in Rivers State provide new and innovative service concept.

Z-Numbers-Based MCDM Approach for Personnel Selection at Institutions of Higher Education for Transportation

Personnel evaluation and selection is an essential part of modern business. Appropriate candidate selection can significantly contribute to companies in terms of increased profit, good culture, reputation, reduced costs, etc. This paper addresses the personnel evaluation and selection problem at the University of Pardubice, Faculty of Transport Engineering (UPCE). Since this is a typical ranking alternative problem where multiple criteria affect the decision, the Z-numbers-based Alternative Ranking Order Method Accounting for the two-step Normalization (AROMAN) is applied. Four Ph.D. candidates are assessed, and the most appropriate is selected to be employed by the UPCE. The Z-numbers fuzzy AROMAN method ranks Ph.D. candidate number four as the most appropriate alternative. To investigate the stability and sensitivity of the Z-numbers fuzzy AROMAN method, the values of parameters β and λ used in the mathematical calculations of the method were changed. The results of sensitivity analysis revealed that the obtained solution is stable. To confirm the robustness of the proposed approach, a comparative analysis is performed. Simple Additive Weighting (SAW), Weighted Product Model (WPM), and Z-number fuzzy TOPSIS were applied. Besides, we applied the fuzzy inferior ratio method as well. The results confirm the high robustness of the proposed Z-numbers fuzzy AROMAN method.

Linear Programming-Based Fuzzy Alternative Ranking Order Method Accounting for Two-Step Normalization for Comprehensive Evaluation of Digital Economy Development in Provincial Regions

As digital technologies continue to reshape economic landscapes, the comprehensive evaluation of digital economy (DE) development in provincial regions becomes a critical endeavor. This article proposes a novel approach, integrating the linear programming method, fuzzy logic, and the alternative ranking order method accounting for two-step normalization (AROMAN), to assess the multifaceted facets of DE growth. The primary contribution of the AROMAN is the coupling of vector and linear normalization techniques in order to produce accurate data structures that are subsequently utilized in calculations. The proposed methodology accommodates the inherent uncertainties and complexities associated with the evaluation process, offering a robust framework for decision-makers. The linear programming aspect optimizes the weightings assigned to different evaluation criteria, ensuring a dynamic and context-specific assessment. By incorporating fuzzy logic, the model captures the vagueness and imprecision inherent in qualitative assessments, providing a more realistic representation of the DE’s multifaceted nature. The AROMAN further refines the ranking process, considering the interdependencies among the criteria and enhancing the accuracy of the evaluation. In order to ascertain the efficacy of the suggested methodology, a case study is undertaken pertaining to provincial areas, showcasing its implementation in the evaluation and a comparison of DE progress in various geographical settings. The outcomes illustrate the capacity of the model to produce perceptive and implementable insights for policymakers, thereby enabling them to make well-informed decisions and implement focused interventions that promote the expansion of the DE. Moreover, managerial implications, theoretical limitations, and a comparative analysis are also given of the proposed method.

Abstract TP77: Stroke Rank Order Correlation but Moderate Absolute Value Drift Between Early Day 2/4 and Day 90 Stroke Outcome Scales: mRS, BI, and NIHSS

Background: In acute stroke trials, the three most commonly used outcome scales are the modified Rankin Scale (mRS) to assess global disability, the Barthel Index (BI) to assess instrumental activities of daily living, and the National Institutes of Health Stroke Scale (NIHSS) to assess neurologic deficit severity. The value of early scale assessments to forecast 3-month disability outcomes has not been formally investigated. Methods: In this cohort study of patients with acute ischemic stroke (AIS) in the NIH Field Administration of Stroke Therapy- Magnesium (FAST-MAG) Phase 3 trial, we assessed the performance of day 2, 4, and 30, BI, and NIHSS to forecast their day 90 counterparts using correlation coefficients and absolute differences. Results: Among 1041 with AIS, mRS on d2 (r=0.67) and d4 (r=0.72) showed strong correlation and mRS on d30 (r=0.89) very strong correlation with d90 mRS. However, value shifts were common. For example, d4 mRS exactly matched d90 mRS in only 35.8%, while 40.2 % had better and 24 % had worse outcomes. D4 BI exactly matched d90 BI in only 41.2%, while 48.5% had better and 11.2% had worse outcomes. For NIHSS, d4 exactly matched d90 value in only 20.5%, while 57.9% had better and 22.6% had worse outcomes. (Figure) Conclusions: mRS, BI, and NIHSS outcomes as early as days 2 and 4 post-stroke correlate strongly with their longterm, 3m outcome rank ordering but only moderately with their absolute values. Delineation of this outcome trajectory provides a foundation for imputing final patient disability, ADL, and deficit outcomes in clinical trials and quality improvement programs.

Co-creation and critical factors for the development of an efficient public e-tourism system

This study identifies the factors that guide the adoption of a public e-tourism system resulting in value co-creation in the UAE. Integrating and comparing factors drawn from the third version of the Technology Acceptance Model (TAM3), the Technology-Task-Fit (TTF) theory, and push-to-use, an Analytic Hierarch Process (AHP) model was implemented with data collected using a structured questionnaire from purposively selected UAE e-tourism experts (N = 15) and analyzed using Microsoft Excel. The findings revealed that usefulness, convenience of use, and push-to-use were the most critical aspects for achieving an efficient public e-tourism system that allows for value co-creation in that order of ranking. The findings also suggest that computer self-efficiency is the most critical factor in effectively establishing an e-tourism system followed by government push-to-use. In conclusion, the findings demonstrate that usefulness and ease-of-use backed by computer self-efficiency, result demonstrability, and output quality are vital for the efficient adoption of a public e-tourism system resulting in value co-creation in the UAE.

Identification of potential pathogenic hepatic super-enhancers regulatory network in high-fat diet induced hyperlipidemia

Hyperlipidemia (HLP) is a prevalent metabolic disorder and a significant risk factor for cardiovascular disease. According to recent discoveries, super-enhancers (SEs) play a role in the increased expression of genes that encode important regulators of both cellular identity and the progression of diseases. However, the underlying function of SEs in the development of HLP is still unknown. We performed an integrative analysis of data on H3K27ac ChIP-seq and RNA sequencing obtained from liver tissues of mice under a low-fat diet (LFD) and high-fat diet (HFD) from GEO database. The rank ordering of super enhancers algorithm was employed for the computation and identification of SEs. A total of 1,877 and 1,847 SEs were identified in the LFD and HFD groups, respectively. The SE inhibitor JQ1 was able to potently reverse lipid deposition and the increased intracellular triglyceride and total cholesterol induced by oleic acid, indicating that SEs are involved in regulating lipid accumulation. Two hundred seventy-eight were considered as HFD-specific SEs (HSEs). GO and KEGG pathway enrichment analysis of the upregulated HSEs-associated genes revealed that they were mainly involved in lipid metabolic pathway. Four hub genes, namely Cd36, Pex11a, Ech1, and Cidec, were identified in the HSEs-associated protein‐protein interaction network, and validated with two other datasets. Finally, we constructed a HSEs-specific regulatory network with Cidec and Cd36 as the core through the prediction and verification of transcription factors. Our study constructed a HSEs-associated regulatory network in the pathogenesis of HLP, providing new ideas for the underlying mechanisms and therapeutic targets of HLP.

Spatiotemporal variation in clam recruitment among beaches in Washington State (USA)

AbstractAbundant clam populations provide ecosystem services and commercial and recreational shellfish harvests worldwide. Two non‐native clams (Manila clams, Ruditapes philippinarum; eastern soft‐shell clams, Mya arenaria) commonly populate mid‐intertidal zones of coarse‐sediment beaches in Washington State, USA. To better understand factors influencing their abundance, we used passive mid‐intertidal spat collectors to determine the density and distribution of clam recruits (150–500 μm shell length) at eight beaches spanning over 500 km and over two reproductive seasons (2016 and 2017). Cumulative seasonal recruitment (clams &lt;500 μm) in 2016 and 2017 at +1.0 m MLLW ranged from 800 to 125,000 m−2 for Ruditapes philippinarum and 50–85,000 m−2 for M. arenaria, and did not differ by substrate type at small scales (gravel vs. crushed shell). At larger scales, different interannual recruitment patterns across sites did not dramatically shift the ranked order of highest and lowest recruitment sites. Cumulative recruitment tended to be lower at warmer sites, while timing of peak recruitment was not consistently related to temperature. Sites also differed in the relative recruitment of the two clam species, but most beaches had greater densities of R. philippinarum than M. arenaria at both recruitment and juvenile stages. Although recruitment densities differed by two orders of magnitude across sites, post‐settlement factors eliminated initial spatial variation within a year, dampening recruitment variability among these beaches that all contribute to commercial and recreational clam production.