International Journal of Uncertainty Fuzziness and Knowledge-Based Systems最新文献

In this global pandemic caused by coronavirus, the role of social media is found to be vital for spreading awareness and faultless news about various aspects of the pandemic. Governments across the world are constantly using available social media platforms to communicate crisis information efficiently to the public, which ultimately making citizens aware about the prevailing conditions. This study systematically investigates how Indian government agencies used social media platform-Twitter to disseminate the relevant information, and to reach out to the citizens during COVID 19 crisis. Spread across various parameters over many days, the twitter data was scrapped from the official Twitter accounts of different government officials. To aggregate and summarize this multi-dimensional data and to process it further, a novel multi-criteria decision making based framework that makes the use of Clustering and Ordered weighted operators is being introduced in this study. Many OWA operators have been introduced in the recent past after their introduction in late 90s, and are used successfully for aggregation purposes in many domains. The summarized values received as an output of the framework are then used to analyze the government response towards COVID 19 situation across various parameters.

Neuro-fuzzy techniques have been widely used in many applications due to their ability to generate interpretable fuzzy rules. Ensemble learning, on the other hand, is an emerging paradigm in artificial intelligence used to improve performance results by combining multiple single learners. This paper aims to develop and evaluate a set of homogeneous ensembles over four medical datasets using hyperparameter tuning of four neuro-fuzzy systems: adaptive neuro-fuzzy inference system (ANFIS), Dynamic evolving neuro-fuzzy system (DENFIS), Hybrid fuzzy inference system (HyFIS), and neuro-fuzzy classifier (NEFCLASS). To address the interpretability challenges and to reduce the complexity of high-dimensional data, the information gain filter was used to identify the most relevant features. After that, the performance of the neuro-fuzzy single learners and ensembles was evaluated using four performance metrics: accuracy, precision, recall, and f1 score. To decide which single learners/ensembles perform better, the Scott-Knott and Borda count techniques were used. The Scott-Knott first groups the models based on the accuracy to find the classifiers appearing in the best cluster, while the Borda count ranks the models based on all the four performance metrics without favoring any of the metrics. Results showed that: (1) The number of the combined single learners positively impacts the performance of the ensembles, (2) Single neuro-fuzzy classifiers demonstrate better or similar performance to the ensembles, but the ensembles still provide better stability of predictions, and (3) Among the ensembles of different models, ANFIS provided the best ensemble results.

The time-varying delay is a peculiar phenomenon that occurs in almost all systems. It can cause numerous problems and instability during system operation. In this paper, the time-varying delay is considered in both the states and input vectors, which is a significant distinction between the proposed method here and previous algorithms. Furthermore, the time-varying delay is unknown but bounded. To address this issue, the Razumikhin approach is applied to the proposed method, as it incorporates a Lyapunov function with the original non-augmented state space of the system models, in contrast to the Krasovskii formula. Moreover, the Razumikhin method performs better and avoids the inherent complexity of the Krasovskii method, particularly when dealing with large delays and disturbances. For achieving output stabilization, the model predictive control (MPC) is designed for the system. The considered system in this paper is an interval type-2 (IT2) fuzzy T-S model, which provides a more accurate estimation of the dynamic model of the system. The online optimization problems are solved using linear matrix inequalities (LMIs), which reduces the computational burden and online computational costs compared to offline and non-LMI approaches. Finally, an example is provided to illustrate the effectiveness of the proposed approach.

In recent years, heterogeneous network/graph representation learning/embedding (HNE) has drawn tremendous attentions from research communities in multiple disciplines. HNE has shown its outstanding performances in various networked data analysis and mining tasks. In fact, most of real-world information networks in multiple fields can be modelled as the heterogeneous information networks (HIN). Thus, the HNE-based techniques can sufficiently capture rich-structured and semantic latent features from the given information network in order to facilitate for different task-driven learning tasks. This is considered as fundamental success of HNE-based approach in comparing with previous traditional homogeneous network/graph based embedding techniques. However, there are recent studies have also demonstrated that the heterogeneous network/graph modelling and embedding through graph neural network (GNN) is not usually reliable. This challenge is original come from the fact that most of real-world heterogeneous networks are considered as incomplete and normally contain a large number of feature noises. Therefore, multiple attempts have proposed recently to overcome this limitation. Within this approach, the meta-path-based heterogeneous graph-structured latent features and GNN-based parameters are jointly learnt and optimized during the embedding process. However, this integrated GNN and heterogeneous graph structure (HGS) learning approach still suffered a challenge of effectively parameterizing and fusing different graph-structured latent features from both GNN- and HGS-based sides into better task-driven friendly and noise-reduced embedding spaces. Therefore, in this paper we proposed a novel attention-supplemented heterogeneous graph structure embedding approach, called as: AGSE. Our proposed AGSE model supports to not only achieve the combined rich heterogeneous structural and GNN-based aggregated node representations but also transform achieved node embeddings into noise-reduced and task-driven friendly embedding space. Extensive experiments in benchmark heterogeneous networked datasets for node classification task showed the effectiveness of our proposed AGSE model in comparing with state-of-the-art network embedding baselines.

Shortest Path Problem (SPP) is mainly used in network optimization; also, it has a wide range of applications such as routing, scheduling, communication and transportation. The main objective of this work is to find the shortest path between two specified nodes by satisfying certain constraints. This modified version of SP is called Constraint Shortest Path (CSP), which establishes a certain limit on selected constraints for the path. The limit for constraint values is precisely specified in traditional CSP problems. But, the precise data may vary due to environmental conditions, traffic and payload. To resolve this, the proposed CSP uses intuitionistic fuzzy numbers to deal with imprecise data. Also, finding an optimal solution in the complex search space of an undirected network is difficult. Hence, Particle Swarm Optimization (PSO) is used in the proposed work to obtain the optimal global solution within feasible regions. A numerical example and the implementation of the proposed work in Matlab 2016a working environment are also illustrated. The simulation analysis shows that the proposed PSO algorithm takes 1.8$$s to find the CSP in a specified undirected network graph, which is comparatively lower than the existing Genetic Algorithm (2.4$$s) and without optimization (5.6$$s).

We present a new classifier that uses a generalized Heronian mean (GHM) operator, and fuzzy robust principal component analysis (FRPCA) algorithms. The similarity classifier was earlier studied with other aggregation operators, including: the ordered weighted averaging (OWA), generalized mean, arithmetic mean among others. Parameters in the GHM operator makes the new classifier suitable for handling a variety of modeling problems involving parameter settings. Motivated by the nature of the GHM operator, we examine which FRPCA algorithm is suitable for use to achieve optimal performance of the new classifier. The effects of dimensionality reduction and fuzziness variable on classification accuracy are examined. The performance of the new classifier is tested on three real-world datasets: fertility, horse-colic, and Haberman’s survival. Compared with previously studied similarity classifiers, the new method achieved improved classification accuracy for the tested datasets. In fertility dataset, the new classifier achieved improvements in accuracy of $14.60\%,19.73\%$, and $23.00\%$ compared with the OWA, generalized mean, and arithmetic mean based classifiers respectively. The new classifier is simpler to implement since it does not require any weight generation criteria as the case is for the OWA based classifier.

The goal of explainable artificial intelligence (XAI) is to solve problems in a way that humans can understand how it does it. For data association there is growing demand for XAI, in which the measurement uncertainty and target (dynamic or/and measurement) model uncertainty are two fundamental problems in maneuvering target tracking in clutter. It commonly suffers of false alarms and missed detections. These situations focus on enhancing explainability, mitigating bias and creating better outcomes for all. Most the probabilistic data association (PDA) methods are weakly able, or even unable, to explain data association. To overcome these situations, the XAI components employed of two modules of Fuzzy-joint probability data association (FJDA) and Fuzzy maneuver compensator (FMC) are first established. Next, these two modules are further employed to construct maneuver tracking scheme, FJDA is then utilized to evaluate the association degree of measurements belonging to different targets and FMC plays compensation role in accordance with maneuver need. The performances of the proposed maneuver tracking scheme were compared with the PDA method and the joint probabilistic data association (JPDA) method using simulated radar surveillance data under a high cluttered environment. The numerical simulation proposed maneuver tracking scheme embedded XAI components FJDA/FCM having a remarkable improvement, due to fully utilize the useful knowledge information in the data association and reduces the impact of measurement uncertainties of the maneuvering target tracking with changing dynamics.

In uncertain information environment, bi-polar preferences can be elicited from experts and processed to be exerted over some weights determination for multiple-agents evaluation. Recently, some weighting methodologies and models in uncertain and preference involved environment with multiple opinions from multiple experts are proposed in some literature. However, in that existing method, when collecting different types of preferences from a single expert, sometimes some subtle cognitive inconsistency may occur. To eliminate such inconsistency, this work elaborately analyzes the possible reasons and proposes some amendment together with a new distinguishable set of formulations for modeling. In addition, we further consider two situations of the weighting models for the problem, with one only considering the situation of single expert with no risk of cognitive inconsistency and the other considering the case of multiple experts wherein some inconsistency might occur. Numerical example and comparison are also presented accordingly.

Multi-Objective Travelling Salesman Problem (MOTSP) is one of the most crucial problems in realistic scenarios, and it is difficult to solve by classical methods. However, it can be solved by evolutionary methods. This paper investigates the Constrained Multi-Objective Travelling Salesman Problem (CMOTSP) and the Constrained Multi-Objective Solid Travelling Salesman Problem (CMOSTSP) under an uncertain environment with zigzag uncertain variables. To solve CMOTSP and CMOSTSP models under uncertain environment, the expected value and optimistic value models are developed using two different ranking criteria of uncertainty theory. The models are transformed to their deterministic forms using the fundamentals of uncertainty. The Models are solved using two solution methodologies Aspiration level-based Multi-Objective Quasi Oppositional Jaya Algorithm (AL-based MOQO Jaya) and Fuzzy Programming Technique (FPT) with linear membership function. Further, the numerical illustration is solved using both methodologies to demonstrate its application. The sensitivity of the OVM model’s objective functions regarding confidence levels is also investigated to look at the variation in the objective function. The paper concludes that the developed approach has solved CMOTSP and CMOSTSP efficiently with an effective output and provides alternative solutions for decision-making to DM.

This paper develops a Fuzzy Adaptive Backstepping Control (FABC) and a Fuzzy Adaptive Backstepping Sliding Mode Control (FABSMC) for Single-Input Single-Output (SISO) nonlinear-systems with unmeasured states. The proposed adaptive schemes are fully compared. Thus, the Fuzzy Type-2 (FT2) concept and the High-Order Integral-Chain Differentiator (HOICD) are used as two universal approximators. Indeed, the first one is employed to approximate the nonlinear system model’s and the second one to estimate the unknown states. Special attention is paid for the used approximators robustness under unmodeled dynamics, parameter variations and process noise.

It should be noted that the asymptotic stability of both the fuzzy adaptive controls and the observer convergence for each scheme have been proved. In addition, the employed schemes have been simulated on a two-tank coupled nonlinear system. Thus, from simulation results, we can prove that the proposed methods guarantee that all signals for closed loop systems are both regular and bounded. Specifically, it can be shown that the performances of the proposed Fuzzy Interval Type-2 (FIT2) schemes are significantly improved compared with Fuzzy Type-1 (FT1) schemes in presence of external disturbances.