Linguistic term sets combining with different types of fuzzy sets to settle decision making problems have been a hot area of decision making research. Linguistic Pythagorean fuzzy sets (LPFSs) and soft sets have strong ability to model uncertainty of decision-making. However, there are still few studies about LPFSs at present and these studies have some deficiencies. Aiming at these problems, we do the following things. Firstly, we give the definition of linguistic Pythagorean fuzzy soft sets (LPFSSs), redefine the entropy for LPFSs and introduce a novel linguistic Pythagorean fuzzy entropy which is more simple and valid. Then we give another way to represent linguistic Pythagorean fuzzy numbers (LPFNs) which can better reflect the characteristics of LPFSs. Based on that, we give the definition of distance measures for LPFSs and propose a series of distance measures for LPFNs and LPFSs. Finally, we improve the TOPSIS method with the proposed entropy and distance measure under LPFSSs environment and then apply the method in two cases. Compared with other methods, it’s shown that our method has better distinguishability in evaluation results and can deal with group decision-making problems under LPFSSs environment.
The theory of intuitionistic fuzzy set has attracted much attention because of its stronger ability in depicting and modelling uncertainty. Recent years, numbers of distance measures of intuitionistic fuzzy sets have been proposed to distinguish the information conveyed by intuitionistic fuzzy sets. However, many existing distance measures cannot meet the requirements of a metric distance. So defining new distance measure for intuitionistic fuzzy sets is important to provide more choice in application. A new method to measure the distance between intuitionistic fuzzy sets is proposed in this paper. The proposed method is developed based on the divergence measure between intuitionistic fuzzy sets, which was firstly introduced by Shannon based on the idea of entropy. The mathematical properties of the new distance measure are discussed, and it is shown that the proposed distance measure satisfies all axiomatic properties of intuitionistic fuzzy distance measure. Numerical examples are presented to verify the effectiveness and reasonability of the new distance measure. Finally, the new distance measure is applied in innovation management to solve the decision making problems. It is illustrated that the proposed distance measure performs better than most of existing measures.
In the conventional federated learning, the local models of multiple clients are trained independently by their privacy data, and the center server generates the shared global model by aggregating local models. However, the global model often fails to adapt to each client due to statistical heterogeneities, such as non-IID data. To address the problem, we propose the Subclass Personalized Federated Learning (SPFL) algorithm for non-IID data. In SPFL, the server uses the Softmax Normalized Gradient Similarity (SNGS) to weight the relationship between clients, and sends the personalized global model to each client. The stage strategy of ResNet is also applied to improve the performance of our algorithm. The experimental results show that the SPFL algorithm used on non-IID data outperforms the vanilla FedAvg, Per-FedAvg, FedUpdate, and pFedMe algorithms, improving the accuracy by 1.81∼18.46% on four datasets (CIFAR10, CIFAR100, MNIST, EMNIST), while still maintaining the state-of-the-art performance on IID data.
The rapid development of meta-learning methods enables the generalized classification of histopathology images with only a handful of new training images. Meta-learning is also named as learning to learn. In this study, we propose a LSTM-model based meta-learning framework for the histopathology image classification. We apply the DoubleOpponent (DO) neurons to model the texture patterns of histopathology images. And the LSTM-model is utilized for the optimization of the meta-learning algorithm to classify the histopathology images. Experiment results on real dataset demonstrated that the proposed method leads in all the measures, namely, recall, precision, F-measure and accuracy.
Objectives Researched the event data recorder and applied to vehicle quality judicial identification. Methods Extensive investigated the application of event data recorder at home and abroad, researched the parameter of event data recorder to apply to accident reconstruction, reappeared the vehicle running condition in the event, and combined with system working theory and design principle to analysis whether there is any fault, and used two cases to illustrate the application of event data recorder. Results The application of event data recorder provided a new identification method for judicial identification. Conclusion The application of event data recorder played an important role in judicial identification and enhanced the science level of judicial identification science.