Preference Matrix Research Articles

Equivalence of plasma and serum for clinical measurement of p-tau217: comparative analyses of four blood-based assays.

Phosphorylated tau (p-tau) 217 is a promising blood biomarker for Alzheimer's disease (AD). However, most p-tau217 assays have been validated solely in ethylenediaminetetraacetic acid (EDTA) plasma, leaving the clinical applicability of serum p-tau217 largely unexplored despite serum being a preferred matrix in many clinical laboratories. To address this gap, we compared p-tau217 concentrations and diagnostic performances in matched plasma and serum samples using four research-use-only assays, including three from commercial sources i.e., Lumipulse, ALZpath, NULISA, and one from University of Pittsburgh. Paired plasma and serum samples were processed from the same venipuncture collection and assessed with the four p-tau217 assays following manufacturer-recommended procedures in two research cohorts (N=84). Plasma and serum p-tau217 levels varied across assays; the ALZpath, Pittsburgh, and NULISA methods showed significantly lower p-tau217 levels in serum compared with plasma (p<0.0001), while Lumipulse showed higher or non-significant differences in serum. Yet, strong correlations (rho >0.8) were observed between plasma and serum p-tau217 pairs. Both plasma and serum p-tau217 demonstrated strong classification accuracies to differentiate clinical AD from normal controls, with high AUC (up to 0.963) for all methods. The exception was the Pittsburgh assay, where plasma p-tau217 had superior AUC than serum p-tau217 (plasma: 0.912, serum: 0.844). The rest of the assays had equivalent accuracies in both matrices. Serum p-tau217 performs equivalently as plasma p-tau217 for most assessed assays. Serum can therefore be used in place of plasma for p-tau217 assessment for research and clinical purposes.

Control of Tissue Strain Is Essential for Enhanced Dermal Innervation in the Three-Dimensional Skin Engineering.

Engineered skin models with sensory innervation are a growing and challenging field of research aimed at applications in regenerative medicine, biosensing, and drug screening. Researchers are attempting to fabricate innervated skin tissues using collagen sponges, cell culture inserts, and microfluidic devices to partially mimic the layered structure of the skin. However, innervation of the full-thickness skin model has not yet been achieved. Here, using the anchoring culture device we previously reported, which is a powerful tool to construct a full-thickness three-dimensional (3D) skin model with balanced tissue contraction forces, we drastically improved dermal layer innervation using a composite hydrogel of collagen and Matrigel (Coll:MG). To determine the preferable hydrogel matrix for neurite extension in the 3D skin construct, DRG neural spheroids were placed at the bottom of the dermal layer composed of various hydrogel scaffold, including type I collagen from different origins (dermis or tendon) and Coll:MG composite hydrogel with different compositions. We showed that the Coll:MG (2:1) composite hydrogel significantly increased vertical neurite extension in the dermal layer, concomitant with the reduced tissue shrinkage during the culture. In contrast, in the collagen-only hydrogel, neurite extension occurred mostly in the horizontal direction, and tissues sometimes detached from the anchors due to significant shrinkage, indicating that tissue shrinkage may affect the direction of neurite extension. To exemplify this idea, 3D skin constructed in the device was partially detached from the anchors to comply with the cell-induced tissue shrinkage and reduce the strain on the tissue. The data showed that the partial allowance of in-plane tissue strain remarkably increased vertical neurite extension compared to the control cultures. Collectively, our results strongly suggest that neurite extension angles can be modulated by adjusting the tissue strain during the culture. Our findings highlight the importance of controlling tissue strain for the advancement of an innervated skin model.

NFGCL: A Negative-sampling-free Graph Contrastive Learning Framework for Recommendation

Recently, Graph Contrastive Learning (GCL) has seen rapid growth in recommender systems. GCL-based Collaborative Filtering (CF) methods typically integrate the primary recommendation task with auxiliary CL tasks, which benefit from two key properties: alignment and uniformity. However, these methods often rely on graph augmentation and negative sampling, which are time-consuming and risk learning incorrect signals. To remedy these issues, we propose a novel Negative-sampling-Free Graph Contrastive Learning (NFGCL) framework to achieve high-quality representations without negative sampling or graph augmentation. Specifically, we use in-batch positive user-item pairs to construct a preference matrix where each element represents cosine similarity between a user and an item. We introduce a novel contrastive objective to make the preference matrix close to an identity matrix, drawing positive pairs closer and distancing negative pairs to achieve better alignment while improving uniformity. Additionally, we propose a simple yet effective representation-level augmentation method that integrates the normalized first-layer Graph Convolutional Networks (GCN) output into the final embeddings, further enhancing alignment. We also employ the uniformity loss to regulate the uniformity of user/item representation distribution. Extensive experiments on three public datasets demonstrate that our method outperforms baseline methods and effectively balances alignment and uniformity.

Review of Aluminium (AA6061) Metal Matrix Composite (MMC) by Stir Casting Method

Industrial development has boosted the use of aluminium to accommodate high demand. Among the most popular aluminium alloys is the AA6061 of the 6xxx series due to it characteristics. AA6061 is a preferred matrix material for Metal Matrix Composite (MMC) fabrication. This study comprehensively reviews an AA6061 aluminium matrix composite produced through the liquid processing technique-stir casting. This review focuses on process parameters, manufacture of conventional mixtures, range of reinforcement, mechanical and tribology properties of AA6061 composites. Significant findings show that increased reinforcement in AA6061 matrix composites greatly improves the properties, such as strength, hardness, ductility, and wear resistance. Producing AA6061 composites with various reinforcements such as organic, inorganic (carbide, oxides, borides and nitrides), hybrid, and other nanomaterials makes the composites more appropriate for high-stress applications. The AA6061 matrix has effectively included a broad spectrum of reinforcing materials, from micro to nanoscale, showcasing the adaptability and versatility of the stir casting method. Future studies should focus on improving AA6061 composite manufacture using advanced casting processes and comprehensive material characterization. These composites have great potential as advanced materials for several industrial applications due to its remarkable mechanical and tribological properties.

Tendons and table salt: A recipe to preserve human DNA

For disaster victim identification, hundreds of samples of human remains have to be extracted from human bodies rapidly for later DNA analysis, often in situation with high humidity, high temperature and no access to cooling facilities. Genetic typing of these samples may be challenging due to degradation of DNA. Until today, bone and teeth are the preferred biological matrix for genetic identification of decomposed human remains. However, processing these samples is time and cost consuming. To address the challenges associated with workload, finances, and sample preservation in the genetic identification of unknown dead bodies, we tested 44 hand and foot tendons from deceased with advanced post mortem changes as alternative DNA material. In addition, we examined commercially available table salt as potential preservation method with 10 tendons over a period of 12 months. The parameters external appearance, DNA quantity and quality have been analysed. Overall, 39 of 44 tendons examined yielded valuable DNA profiles with at least ten autosomal STR loci. Storage in table salt not only increased DNA yield but also resulted in high quality STR profiles of the preserved tendons throughout the study period of 12 months compared to storage without salt. According to the results of this study, we recommend hand and foot tendons as alternative DNA material from bodies with advanced post mortem changes. We particularly recommend commercially available table salt as a storage medium for the preservation of DNA samples. Under field conditions without cooling facilities and the need to store samples prior to DNA analysis table salt is a good storage medium: it preserves DNA and is easily available.

Managing flexible linguistic expressions with subjective preferences and objective information in group decision-making: A perspective based on personalized individual semantics

Flexible linguistic expression (FLE) is a complex linguistic representation that almost generalizes all distributed representations, posing the following challenges for the personalized individual semantics (PIS) issues in group decision-making (GDM) with FLEs. (1) The available measurement methods for FLEs are limited and suffer from some unreasonable or poorly discriminated deficiencies. (2) The existing research on FLEs has failed to fully utilize decision information incorporating subjective preferences and objective information, and the format of FLEs may be altered when optimizing PISs. To address these challenges, this paper focuses on the GDM problems under subjective preferences and objective information, both of which are characterized by FLEs, and introduces a novel FLE-based GDM method with PIS. Firstly, a new similarity measure for FLEs is proposed, considering both external and internal characteristics, namely semantics, elements, and symbolic proportions of linguistic term subsets. Secondly, by utilizing the similarity between the decision-maker's preference matrix and decision matrix, as well as the consistency of the preference matrix, a PIS optimization model is constructed to determine the personalized semantics of individual decision-makers. Subsequently, an optimization model is constructed, focused on maximizing the consensus level to determine the differentiated weights of decision-makers. Furthermore, individual evaluation information is aggregated into group evaluation information, a personalized semantic model is constructed to determine the personalized semantics of the group, and the selection process for alternatives is implemented. Finally, the effectiveness and advantages of our method are illustrated through its application in airline service quality assessment and by comparing it with alternative approaches.

MOOCs video recommendation using low-rank and sparse matrix factorization with inter-entity relations and intra-entity affinity information

PurposeThe serious information overload problem of MOOCs videos decreases the learning efficiency of the students and the utilization rate of the videos. There are two problems worthy of attention for the matrix factorization (MF)-based video learning resource recommender systems. Those methods suffer from the sparsity problem of the user-item rating matrix, while side information about user and item is seldom used to guide the learning procedure of the MF. MethodTo address those two problems, we proposed a new MOOCs video resource recommender LSMFERLI based on Low-rank and Sparse Matrix Factorization (LSMF) with the guidance of the inter-Entity Relations and intra-entity Latent Information of the students and videos. Firstly, we construct the inter-entity relation matrices and intra-entity latent preference matrix for the students. Secondly, we construct the inter-entity relation matrices and intra-entity affinity matrix for the videos. Lastly, with the guidance of the inter-entity relation and intra-entity affinity matrices of the students and videos, the student-video rating matrix is factorized into a low-rank matrix and a sparse matrix by the alternative iteration optimization scheme. ConclusionsExperimental results on dataset MOOCcube indicate that LSMFERLI outperforms 7 state-of-the-art methods in terms of the HR@K and NDCG@K(K = 5,10,15) indicators increased by an average of 20.6 % and 21.0 %, respectively.

Exploring the dose-response of landscape preference: A case study in Singapore

Despite pursuing greening initiatives, cities might experience excessive greenery at the neighborhood level. The intension of this research is to promote ecosystem services and avoid the mantle of so-called “green urbanism”. Hence, examining the dose-response relationship is essential to understanding the intricate interplay between residential greenery quantity and landscape preference. Drawing upon preference assessment data from multiple neighborhoods with 30 typical scenes (n = 147), we employ Kaplan's landscape preference matrix (KLPM) to quantify the spatial distribution of greenery and individuals' preference assessment from the perspectives of Mystery, Coherence, Legibility, and Complexity. Results from multiple and ridge regressions indicate that the dose-response curve follows a polynomial shape, with landscape preferences reaching a threshold as the NDVI value increased from 0.35 to 0.42. Beyond the threshold, the trend of landscape preference levels tapers off and eventually plateaus. Therefore, the conclusion of our cross-sectional experiment adequately reflects the threshold effect of landscape preference toward greenery quantity. Furthermore, the weight of Mystery, Coherence, and Legibility in KLPM all showed a positive strong correlation, but the dose-response relationship cannot be fully explained by Complexity. These results inform how designers can more effectively promote optimal doses of nearby nature to circumvent worldwide overheated "Garden City” initiatives.

Evaluate the competence of fuzzy preference using statistical conflict in the frame of DS evidence theory

Fuzzy preference based DS evidence theory effectively solves the combined explosion ascribing to the increasing of the cardinality of the frame of discernment in the application. However, limited work has focused on evaluating the performance of the binary classifier (i.e. dynamic improvement of the according fuzzy preference element). The conflict coefficient can well reflect the relationship between various pieces of evidence and effectively evaluate the ability of each classifier. By statistically analyzing prior conflict information, the conflict interval range is determined to more effectively correct the fuzzy preference matrix. Therefore, we proposed a new method to evaluate the competence of fuzzy preference using statistical conflict in the frame of DS evidence theory. First, a binary transformation of the multiclassification problem is performed to generate the BPA and the matrix was generated by orthogonal fusion of multiple BPAs. Second, calculate the conflict coefficients and make statistics. The statistical conflict coefficients are combined and the conflict interval is solved to modify the construction of the fuzzy preference matrix. Third, based on the measured data in the feature space, the matrix is constructed by majority voting using the k-nearest neighbors of correctly classified objects. Finally, the two matrices are fused by determining the relationship between the conflict coefficient and the threshold value of each binary pair. The non-dominance degree is calculated based on the modified fuzzy preference matrix to make the final decision. Analysis and results of application examples and several datasets in practical application are used to demonstrate effectiveness and flexibility of the proposed method.

An Incentive Mechanism-Based Minimum Adjustment Consensus Model Under Dynamic Trust Relationship.

In traditional group decision making, the inconsistent experts are usually forced to make compromises toward the group opinion to increase the group consensus level. However, the strategy of reaching group consensus via an incentive mechanism encouraging adjustment of preferences is more effective than forcing, which is the aim of this article. Specifically, this article establishes a novel incentive mechanism to support group consensus under dynamic trust relationship. First, the supremum and infimum incentives-based rule driven by trust relationship is defined. Based on the assumption that if incentive conditions are met, then experts will be willing to adjust their preferences, the incentive behavior-driven minimum adjustment consensus model is developed to generate optimal incentive-based recommendation preferences. Thus, the proposed incentive mechanism can effectively reduce the preference adjustment cost and promote group consensus reaching. Third, the updated trust relationships between experts are shown to be strengthen by the proposed incentive-driven preference revision. Consequently, the optimization model based on trust interaction relationship is constructed to obtain the final group preference matrix. Finally, a supplier selection case of high-end medical equipment is provided to illustrate the proposed method and show the rationality and advantages of the proposed methodology with both a sensitivity analysis and a comparison analysis.

A two-stage stochastic preference analysis on group decision making with interval additive preference relations

Interval additive preference relation (IAPR) is an efficient tool for describing the uncertain pairwise comparisons between alternatives in group decision-making (GDM). Existing GDM methods with IAPRs view the preference value as a whole and establish concrete conflict/consensus mechanisms for fusing individual opinions while neglecting the stochastic preference and aggregation modes to meet the different demands of decision. Thus, to reflect the stochastic characteristic in both preference representation and group preference aggregation, this paper presents a novel two-stage stochastic preference analysis on GDM with IAPRs. In first stage, the method involves extracting stochastic additive preference relations (SAPRs) from the IAPRs. A preference relations matrix is then established using a logarithmic least squares model applied to the SAPRs. In second stage, the weighted preference matrix is synthesized by combining the preference relations matrix with the decision makers’ weight space. This leads to the computation of positive and negative ideal preference solutions, as well as positive and negative preference distances. Subsequently, the relative preference closeness is calculated based on these distances. Relative ranking results are determined by an algorithm evaluating pairwise comparisons among alternatives, resulting in transitional preference rankings. The second stage involve deriving ranking acceptance indices and ranking confidence degrees through a continuous extraction process, facilitating the establishment of the ultimate best preference rankings. The proposed method’s effectiveness and validity are demonstrated by numerical examples and Monte Carlo simulations.

Healable and multi driving mode soft actuator enabled by disulfide-bonded liquid crystal elastomers

Due to the great potential in many applications, multi-responsive soft actuators as a kind of smart materials have attracted great interest from the scientific community. Liquid crystal elastomers (LCEs) are gradually becoming their preferred candidate matrix because of their unique mesogenic anisotropy and polymer elasticity. Here, we designed and synthesized a new disulfide-bonded aromatic molecule (named DSAC), which was used to endow the LCE matrix with a self-healing effect to extend the service life. The large conjugate structure greatly reduces the bond energy of the disulfide bond, thus allowing self-healing to occur at a lower temperature. Furthermore, the LCE matrix was coated with a novel type of functional layer with photothermal and conductive properties to make it be driven by multi-modes. Based on the above, we have prepared a healable and dual-driving mode soft actuator enabled by disulfide-bonded liquid crystal elastomers with multiple stimulus–response properties, which shows great potential in flexible electronics, artificial muscles and soft robots.

Enantiomer-specific analysis of amphetamine in urine, oral fluid and blood.

Illegal amphetamine is usually composed of a racemic mixture of the two enantiomers (S)- and (R)-amphetamine. However, when amphetamine is used in medical treatment, the more potent (S)-amphetamine enantiomer is used. Enantiomer-specific analysis of (S)- and (R)-amphetamine is therefore used to separate legal medical use from illegal recreational use. The aim of the present study was to describe our experience with enantiomer-specific analysis of amphetamine in urine and oral fluid, as well as blood, and examine whether the distribution of the two enantiomers seems to be the same in different matrices. We investigated 1,722 urine samples and 1,977 oral fluid samples from prison inmates, and 652 blood samples from suspected drugged drivers, where prescription of amphetamine was reported. Analyses were performed using ultra high performance liquid chromatography-tandem mass spectrometry (UHPLC-MS-MS). The enantiomer separation was achieved by using a chiral column, and results from the method validation are reported. Samples containing <60% (S)-amphetamine were interpreted as representing illegal use of amphetamine. The distribution of the two enantiomers was compared between different matrices. In urine and oral fluid, the mean amount of (S)-amphetamine was 45.2 and 43.7%, respectively, while in blood, the mean amount of (S)-amphetamine was 45.8%. There was no statistically significant difference in the amount of (S)-amphetamine between urine and oral fluid samples and between urine and blood samples, but the difference was significant in blood compared to oral fluid samples (P < 0.001). Comparison of urine and oral fluid between similar populations indicated that enantiomers of amphetamine can be interpreted in the same way, although marginally higher amounts of (R)-amphetamine may occur in oral fluid. Oral fluid, having several advantages, especially during collection, could be a preferred matrix in testing for illegal amphetamine intake in users of medical amphetamine.

Medicinal ethnoveterinary plants used for treating livestock ailments in the omo-gibe and rift valley basins of Ethiopia

BackgroundTraditional medical practices have been used to maintain animal health for millennia and have been passed down orally from generation to generation. In Ethiopia also, plants are the primary means by which the indigenous people in remote areas treat the illnesses of their animals. The present study was therefore, carried out to document the type and distribution of medicinal plants of the county.MethodsTo collect ethnobotanical information, a total of 205 informants (133 men and 72 women) were selected. Among these 121 traditional medicine practitioners, while the remaining 84 were selected through a systematic random sampling method. Ethnobotanical data were collected between January 2023 and August 2023 through semi-structured interviews, participant observation, guided filed walks and focus group discussions. Using descriptive statistics, the ethnobotanical data were analyzed for the Informant Consensus Factor (ICF) and Fidelity Level (FL) values, preference, and direct matrix rankings. SPSS 26.0 and PAST 4.11 software were used in data analysis.ResultsTotally, 78 ethnoveterinary medicinal plants distributed in 36 families were identified in the study area. Asteraceae was the dominant family with 9 species (14%), followed by Euphorbiaceae with 8 species (12%). Herbs 42(56%), wild collected 62 (66%), and leaf part (52%) made the highest share of the plant species. Hordeum vulgare L. had the highest fidelity level (FL = 98%) for treating bone fractures. Blackleg, bloat, and endoparsistes each had the highest values of the consensus factor among the informants (ICF = 1). According to preference ranking, Withania somnifera was the most potent remedy for treating blackleg. Knowledge of medicinal plants was shared through storytelling within families.ConclusionIn the study area, there is broad access to traditional medicinal plants that can treat ailments in animals. Conservation efforts should be prioritized to protect medicinal plants from threats such as agricultural expansion, drought, and development. Further research should be conducted to explore the potential of different medicinal plants for treating common livestock ailments.

A group decision-making method for technology selection of petrochemical enterprises

Selecting appropriate energy conservation and emission reduction technologies is essential for the sustainable developments of petrochemical enterprises. It usually involves experienced experts from multi-domain and thus can be regarded as a large-scale group decision-making problem. Existing methods assumed that different experts use the same linguistic term set following the same semantics to express their evaluation values, and derived the ranking of alternatives by an indirectly obtained group preference matrix. Such an approach may induce information loss. To avoid these defects, this study introduces a model based on rank centrality and expert differences to choose appropriate technologies for petrochemical enterprises. First, an attitude parameter is introduced into a linguistic scale function to unify different judgment benchmarks of experts regarding the same linguistic term. Next, a direct weight elicitation method based on rank centrality is developed to deduce the individual rank vector of alternatives. A grey clustering method based on the similarities of individual rank vectors is extended to cluster experts into small groups, and the silhouette coefficient is used to determine clusters weights. Afterwards, the rank vectors of alternatives corresponding to subgroups and the global rank vector are calculated by the weighted averaging operator. Alternatives are ranked by the rank vector of the global group. Finally, four technologies were evaluated by the model, and the preferred technology was identified. The findings indicate that the proposed technology selection model alleviates information loss, elevates group consensus level, and provides a perspective for petrochemical enterprises in search of energy conservation and emission reduction technologies.

5G new radio multiuser precoding in geostationary satellite networks

SummaryThe multiuser precoding procedures standardized by the 3rd Generation Partnership Project (3GPP) for new radio (NR) represent a new opportunity to realize satellite precoding for capacity improvement, given the integration of non‐terrestrial networks in the 3GPP framework. This work analyses and numerically evaluates them for the geostationary satellite scenarios consolidated by the 3GPP. It shows that the preferred precoding matrix indicated by NR codebooks is not able to handle properly the inter‐beam interference and should only be used to estimate the propagation channel. To this aim, we propose the use of modified type II port selection codebooks reporting on the strongest beams, which correspond to the higher interferences. Since the obtention of high precoding gains requires the operation in rather high signal‐to‐noise regimes and the proper estimation of the interfering contributions from many beams, we propose to enlarge the maximum number of beams reported by each precoding matrix indicator (PMI) report, currently limited to four. As alternative, we also evaluate a sequential multi‐report framework considering the effects of channel‐state information aging and the user equipment (UE) processing times. It preserves the current limits on the number of reported beams but increases the system overhead signalling. The quantization of estimated coefficients is also analysed, not only from the overhead point of view but also to properly report weak interferers corresponding to distant beams, due to the increase in the number of reported beams. Moreover, we identify the minimum interference power a UE should be able to detect to achieve the theoretical large precoding gains.

Понятийный аппарат и практика оценки и контроля налоговых рисков субъектов предпринимательства в условиях изменчивости налоговой среды

Tax risks are situations in which an enterprise may face loss or damage related to taxation as a result of incorrect application of tax rules, incorrect interpretation of legislation or other circumstances that may lead to negative financial consequences. Enterprises often suffer losses either due to tax due to the accountant's ignorance of tax legislation, or due to the excessive riskiness of the accountant's decision-making and subsequent accrual of penalties, fines, additional payments and even litigation. However, methodological foundations and methodological apparatus for assessing tax risks of business entities have not yet been fully studied. It is accurately proved that study of tax risks classification contributes to more accurate accounting, analysis and management of financial risks of an enterprise, as well as helps to comply with laws and regulations, which is a key factor for successful and sustainable operation of any company. There is an extensive classification of tax risks according to different principles: by type of tax risk, by the source of occurrence (by risk determining factors), according to the time of occurrence, according to the amount of possible losses, according to the type of tax risks consequences, according to the object of connection with other types of risks, according to the nature of consequences, according to the possibility to overcome, and according to the level of probability of tax risks occurrence. Tax risk assessment is an important and integral part of the enterprises tax management in the Russian Federation. This process allows business entities identifying and analyzing potential threats and negative consequences related to taxation, and develop strategies to minimize and manage them. The purpose of the study is to analyze tax risks which enterprises in the Russian Federation face and to develop methods and recommendations for their assessment and management in order to ensure financial stability and compliance with tax legislation. The methodology of the study is formed of standardized methods to assess the degree of compliance of the enterprise tax practice with applicable laws and standards. Empirical methods, in particular the analysis of statistical data, and case studies are used. As a result, the key factors influencing tax risks occurrence are highlighted. Modern methods of assessing tax risks matrix are summarized and practice of applying of tax risk analysis matrix and matrix of paired preferences of tax risks is presented. Ways to minimize tax risks were proposed using the example of LLC Krasnaya Polyana Agro-Industrial Complex, and a scheme of its tax risk control system was proposed. The results of tax risk assessment serve the basis for the development of tax strategies and risk management processes. Enterprises can minimize tax risks by optimizing the tax strategy proposed by the authors, creating reserves and constantly monitoring tax environment changes.

Double stochastic preference analysis on group decision making with interval additive preference relations

Interval additive preference relations (IAPRs) provide a valuable framework for representing uncertain pairwise comparisons among alternatives in group decision-making (GDM). Conventional GDM methods that incorporate IAPRs often treat preference values as uniform entities and establish dedicated conflict or consensus mechanisms for amalgamating individual opinions while overlooking the stochastic preferences and lacking the flexibility to address diverse decision requirements. Therefore, to account for the stochastic characteristics in both preference representation and group preference aggregation, this paper introduces an innovative double stochastic preference analysis for GDM with IAPRs. In the initial phase, stochastic additive preference relations are extracted from IAPRs using a specific density function. A best preference matrix is then constructed through a logarithmic consistency minimum deviation model. In the second phase, preference comparison intervals are synthesized to encapsulate the collective viewpoints of decision-makers. Subsequently, ranking acceptability indices and preference confidence degrees are computed through processing of the preference ranking judgment space, leading to the derivation of the final preference rankings. The experimental results, obtained through Monte Carlo simulations for a site selection case in Anhui Province, China, demonstrate the effectiveness and validity of the proposed method. The ranking results achieved a confidence degree exceeding 89%, offering insights into decision-makers’ weight distribution for various ranking scenarios.

On incomplete matrix information completion methods and opinion evolution: Matrix factorization towards adjacency preferences

Completeness and consistency are widely acknowledged as indispensable prerequisites for applying fuzzy preference relations in resolving real-world problems. Interactions between decision-makers in the field of group decision-making can have a profound effect on preference changes, especially in situations of incomplete information. However, preference matrices generated using existing models often suffer from low consistency during opinion evolution, which further increases the risk of wrong decisions. To cope with these problems, this paper proposes an opinion evolution method under incomplete information. The method consists of two main components: (1) A preference prediction model based on matrix factorization algorithm and social trust networks under incomplete information, and (2) A model for decision-makers’ preference evolution under dynamic trust scenarios. In Part (1), to simplify the preference collection method and reduce the occurrence of inconsistent information, this paper presents the concept of a consistent adjacency fuzzy preference matrix. To advance the utilization of machine learning algorithms for consistent preference relations, this paper employs the matrix factorization method with stochastic gradient descent to predict missing preferences in adjacency fuzzy preference matrix. In Part (2), to address the shortcomings in past opinion evolution models, this paper simulates the change of decision-makers’ opinions in different stages of interactions under variable trust weights by using alternative ranking similarity and trust thresholds. To illustrate the practical application of the proposed model, this article uses the enterprise supplier selection problem as a case study. The model’s effectiveness and feasibility are then validated through a comparative analysis with existing methods.

Studying the API Distribution of Controlled Release Formulations Produced via Continuous Twin-Screw Wet Granulation: Influence of Matrix Former, Filler and Process Parameters.

Hydroxypropyl methylcellulose (HPMC) is a preferred hydrophilic matrix former for controlled release formulations produced through continuous twin-screw wet granulation. However, a non-homogeneous API distribution over sieve fractions with underdosing in the fines fraction (<150 µm) was previously reported. This could result in content uniformity issues during downstream processing. Therefore, the current study investigated the root cause of the non-homogeneous theophylline distribution. The effect of process parameters (L/S-ratio and screw configuration) and formulation parameters (matrix former and filler type) on content uniformity was studied. Next, the influence of the formulation parameters on tableting and dissolution behavior was investigated. Altering the L/S-ratio or using a more aggressive screw configuration did not result in a homogeneous API distribution over the granule sieve fractions. Using microcrystalline cellulose (MCC) as filler improved the API distribution due to its similar behavior as HPMC. As excluding HPMC or including a hydrophobic matrix former (Kollidon SR) yielded granules with a homogeneous API distribution, HPMC was identified as the root cause of the non-homogeneous API distribution. This was linked to its fast hydration and swelling (irrespective of the HPMC grade) upon addition of the granulation liquid.