The probabilistic linguistic term set (PLTS), composed by linguistic terms and their probabilities, is effective to represent uncertain evaluations. Considering that interval probability is more powerful than the precise form in describing uncertainty, this study introduces the PLTS with interval probabilities. Based on belief and plausibility measures, in this article, we discuss how to translate complex qualitative evaluations, which may be composed by both interval probabilities and interval linguistic terms, to the PLTS with interval probabilities. Utility-based translation approaches are proposed, which further shows the ability of the PLTS with interval probabilities in representing quantitative information. In addition, a probabilistic linguistic dominance method is developed to compare PLTSs. Integrating optimization models with the Dempster–Shafer theory, we present an aggregation method to estimate the maximum and minimum PLTSs obtained from the combination. Furthermore, a multicriteria decision-making method is introduced considering both the comprehensive evaluations of alternatives and the ability to achieve the tolerance and expectation values of criteria. The applicability of the proposed approach is illustrated by a case study of shelter selection.
Classical clustering algorithms such as k-means face limitations in handling clusters with heterogeneous shapes, densities, and sizes, while exhibiting sensitivity to initial centroid selection. To overcome these challenges, this article proposes a novel clustering framework based on regenerated granular ball (RGGB) with the principle of justifiable granularity. Unlike existing granular-ball (GB) techniques that overemphasize purity criteria at the expense of uncontrolled ball sizes, RGGB dynamically adjusts granularity levels through iterative regeneration, achieving an optimal balance between detailed data representation and computational efficiency. This adaptability enhances stability in capturing data similarities while mitigating sensitivity to initialization. To validate the method, we integrate RGGB with a novel k-nearest neighbor (KNN) classifier using regenerated GBs to evaluate classification performance and demonstrate practical applications. Experiments on diverse public and realistic datasets demonstrate that the RGGB-based KNN algorithm consistently outperforms existing techniques, including traditional KNN and other methods, making a promising advancement in clustering and classification tasks.
Automatic emotion recognition plays a key role in human-computer interactions. Multimodal emotion recognition has attracted much attention in recent years. When multimodalities are used, different modalities interact with each other and the obtained results tend to be accurate in general. However, there are also cases of unimodal anomalies. Most of the existing studies do not take into account the existence of outliers in the multimodality, which leads to low accuracy of the prediction results. This paper proposes fuzzy weighted support vector machine for regression (FWSVR) to deal with outliers and prediction errors. We design an automatic affective recognition model structure to analyze continuous dimension emotions based on multimodality (audio and visual). The LIRIS-ACCEDE database is used in this work. Experimental results indicate that the concordance correlation coefficient (CCC) is 0.9456 for arousal and 0.9183 for valence on the test set. The fusion result obtained when using fuzzy weighting is much better than the direct fusion one.
Running evolutionary algorithms in parallel is an intuitive way to speed up the process of solving large-scale multi-objective optimization problems, which have hundreds or thousands of decision variables. However, the framework of the existing multi-objective evolutionary algorithms seriously limits their parallelization. During each iteration, the environmental selection operators present in the existing framework need to collect and compare all the candidate solutions to balance the convergence and diversity, thus dividing the whole evolutionary process into a series of dependent sub-processes and resulting in frequent data transmission. To address this issue, we propose a novel parallel framework that separates the environmental selection operator from the entire evolutionary process, evidently removing the dependencies among sub-processes and reducing the data transmission. On the basis of the parallel framework, a new parallel evolutionary algorithm, namely PEA, is designed. In PEA, the convergence is achieved by a series of independent sub-populations, and the diversity is merely emphasized at the converged solutions from each subpopulation, which is helpful for avoiding that the environmental selection operator limits the parallelization of the algorithm. Moreover, a new environmental selection strategy is proposed to improve the diversity without considering the convergence. To assess the performance of the proposed PEA, we compare it with five representative multi-objective evolutionary algorithms in terms of both the convergence and diversity. The performance of the parallel framework is also analyzed by comparing with two existing parallel models. The experimental results demonstrate the superiority of the proposed parallel algorithms in terms of the convergence, diversity, and speedup.
