Reimann, Jan NiclasSchwung, AndreasDing, Steven X.
17页
查看更多>>摘要：Deep neural networks (DNN) are mainly black boxes, generally suffering from bad interpretability of their behavior and the results obtained. Hence, a human can not easily derive the relations modeled by the network. A reasonable way to provide interpretability for humans are logical rules. In this paper we propose neural logic rule layers (NLRL), which are able to represent arbitrary logic rules in terms of their conjunctive and disjunctive normal forms. Stacking various layers, we are theoretically able to represent arbitrary complex rules by the resulting neural network architecture. The NLRL are end-to-end trainable allowing to learn logic rules directly on the given data without needing any background information about the origin. We show in experiments, that NLRL-enhanced neural networks can model arithmetic and logical operations over the input values. Furthermore, we apply NLRL to image classification tasks and show that interpretability is provided without sacrificing classification performance by exchanging the fully-connected head of the network. We also apply NLRL to a real world industrial control problem where the task is to model the discrete control behaviour of a programmable logic controller (PLC), following a basic step sequence.(c) 2022 Elsevier Inc. All rights reserved.
Dignos, AntonGamper, JohannBlumenthal, David B.Bougleux, Sebastien...
20页
查看更多>>摘要：Matchings between objects from two datasets, domains, or ontologies have to be computed in various application scenarios. One often used meta-approach -which we call bipartite data matching-is to leverage domain knowledge for defining costs between the objects that should be matched, and to then use the classical Hungarian algorithm to compute a minimum cost bipartite matching. In this paper, we introduce and study the problem of enumerating K dissimilar minimum cost bipartite matchings. We formalize this problem, prove that it is NP-hard, and present heuristics based on greedy dynamic programming. The presented enumeration techniques are not only interesting in themselves, but also mitigate an often overlooked shortcoming of bipartite data matching, namely, that it is sensitive w. r. t. the storage order of the input data. Extensive experiments show that our enumeration heuristics clearly outperform existing algorithms in terms of dissimilarity of the obtained matchings, that they are effective at rendering bipartite data matching approaches more robust w. r. t. random storage order, and that they significantly improve the upper bounds of state-of-the art algorithms for graph edit distance computation that are based on bipartite data matching.(c) 2022 Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
查看更多>>摘要：In recent years, with the development of Internet technology, recommender systems have been widely used by virtue of their ability to meet the personalized needs of users. In order to make full use of users' interactive behaviors, session-based recommender systems have attracted growing research interest. In previous session-based recommender systems, users' historical interactive behavior is utilized to train and update users' preferences, but users' responses to the current recommendation results (immediate feedback) are not effectively exploited to optimize the recommendation strategy. This leads to the decrease of subsequent recommendation accuracy. Aiming at this problem, based on the Recurrent Neural Network (RNN), this paper combines Reinforcement Learning (RL) and Generative Adversarial Networks (GANs). We fully exploit the users' immediate feedback with RL, and simultaneously take advantage of GAN to satisfy the requirement of training data brought by RL. Furthermore, we optimize the negative sampling method and propose Deep Generative Adversarial Networks-based Collaborative Filtering (DCFGAN). The experimental results show that this algorithm can effectively improve the recommendation accuracy in session-based recommender systems. (c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：In recent years, many different membrane-inspired evolutionary algorithms have been proposed to solve various complex optimization problems. Considering membrane systems' powerful computing performance and parallel capability, it has outstanding potential in solving multi-task optimization problems. However, there is no research to explore the performance of membrane-inspired evolutionary algorithms in solving multi-task optimization problems. In this paper, a novel membrane-inspired evolutionary framework with a hybrid dynamic membrane structure is proposed to solve the multi-objective multi-task optimization problems. First, a novel membrane-inspired two-stage evolution strategy algorithm is proposed as the algorithm in the membrane to improve the convergence of the algorithm and the diversity of multisets. Second, the information molecule concentration vector is proposed to reduce negative information transfer. The information molecule concentration vector is inspired by the binding process of information molecules and receptors and can control the information transfer probability adaptively. Finally, comprehensive experimental results show that the proposed algorithm performs better than most advanced multi-objective evolutionary multitasking algorithms. (c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：An approach based on forced Duffing Equation with cubic term is proposed to solve the single risk factor control problem in complex social-technical systems from the micro perspective. The single risk control problem is transformed into a system vibration control problem. Firstly, the relationships among Duffing Equation, system state sudden change and single risk factor control are analyzed. System control point crosses the bifurcation set is an important condition for the system state sudden change, the Duffing Equation can be used to establish the vibration equation of the complex social-technical system, and the risk control method of the complex social-technical system can be deduced based on the vibration equation. Secondly, the interference of the external environment in Duffing Equation is defined as the energy change of risk pulse. Next, we establish the system risk control equation and derive its bifurcation response equation with stable solutions. Under three safety protection measures conditions, including strengthening internal damping of system, reducing external excitation influence, strengthening internal damping of the system while reducing external excitation influence, three system risk control strengthening equations are established and their corresponding bifurcation response equations with stable solutions are derived, respectively. Finally, a real-world case study is conducted.(c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：Deep learning based soft analyzers are important for modern industrial process monitoring and measurement, which aim to establish prediction models between quality data and easy-to-measure variables. However, in traditional deep learning methods, the guidance of quality information on feature extraction is insufficient and easily reduces as data dimension increases. In this paper, a stacked maximal quality-driven autoencoder (SMQAE) is proposed to extract maximal quality-relevant features for soft analyzers. In each maximal quality-driven autoencoder, quality variables are reconstructed together with the input variables in the output layer. The SMQAE ensures that the influence of the quality part and input part on the reconstruction are the same. And the maximal information coefficient (MIC), which is not limited to any specific function type, is exploited to enhance the importance of quality-related variables in the input part. With the constraint of the quality equivalence strategy and variable importance evaluation based on MIC, the SMQAE maximizes the guidance of the quality variables during feature learning without the interference of the data dimensions. Therefore, the SMQAE can extract quality relevant features for complex high-dimensional data. The rationality, superiority and robustness of SMQAE based soft analyzers are validated on four simulated scenarios and two industrial processes.(c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：Information security and image encryption have rapidly developed in recent years. This study presents an iterative method for vectors with fractal characteristics. Based on the idea of combining fractal and algebraic theories, the double parameters fractal sorting vector (DPFSV) was originally proposed, which incorporates features such as complexity, self similarity, and iteration. Based on the practical application requirements, the DPFSV has a good regulatory vector length and increases the diversity of information location change laws. Hence, this study uses the DPFSV to control iterative node relationships in spatiotemporal chaotic systems. The new spatiotemporal chaotic system proposed based on DPFSV exhibits better dynamics than that of the coupled map lattice (CML) system, verified by comparing their Kolmogorov-Sinai entropy, bifurcation diagram, information entropy, and mutual information. Therefore, DPFSV-CML is more suitable in cryptography than CML. Combining the DPFSV and spatiotemporal chaotic system, this study constructs a new cryptographic system to implement permutation-diffusion synchronous encryption. This new encryption method has a good encryption effect and is resistant to various attacks. The security of this proposed cryptographic system is verified through various security analyses and simulated attack validations.(c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：In this paper, we propose a Two-stage Differential Evolution (TDE) with novel parameter control for real parameter single objective global optimization. In the TDE algorithm, the whole evolution is divided into two stages and each stage employs a unique mutation strategy. The mutation strategy in the first stage is a novel historical-solution based mutation strategy, which can get better perception of the landscape of the objective; the mutation strategy in the second stage is an inferior-solution based mutation strategy, which can maintain better diversity of trial vector candidates while keeping better convergence speed. Furthermore, the parameter control of our TDE is novel, which means that these adaptations of control parameters are different from those in the literature: First, the adaptation schemes both for scale factor F and crossover rate CR are fitness-independent. Second, different from the fixed population size and the gradually reduced population size, the population adaptation in TDE has two different stages. Third, a stagnation indicator is proposed and a population enhancement technique can be launched if necessary when a certain individual is in the stagnation status. We examine the TDE algorithm under a relative large number of benchmarks from CEC2013, CEC2014 and CEC2017 test suites for real-parameter single objective global optimization, and the experiment results show the competitiveness of our TDE algorithm with several recently proposed state-of-the-art DE variants, e.g. it obtained 20 similar or better performance improvements out of the total 30 benchmarks in comparison with the winner algorithm, the LSHADE algorithm, of the CEC2014 competition and it also obtained 19 similar or better performance improvements out of the total 30 benchmarks in comparison with the winner DE variant, the jSO algorithm, of the CEC2017 competition. (c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：For data analysis, learning treatment rules in stratified medicine require the optimization of multiple responses. A common approach is to use a multi-objective function to find the optimal setting of the controllable factors. For patients, the optimal setting is a treatment regimen that yields the optimal value of potential responses. However, subclasses of patients are often stratified by their covariates. Thus, this paper proposes a new model called constrained optimization for stratified treatment rules (COSTAR) with multiple responses. This model incorporates covariates to build separate models for optimal responses and stratifies the patients with the balancing score from covariates. The optimal solution enables us to choose the optimal treatment for each subclass of patients. Theoretical results guarantee the identifiability of the solutions with conditional optimal values of multiple responses from survival probabilities. Examples of experiments with factorial designs and survival data validate the efficacy of the proposed method. The results suggest that this method improves the significance of the parameters and the adjusted R-2 in fitting on the primary response, while the unsupervised clustering method (i.e., k-means) does not. This method, with the fitting model, is more interpretable than the conventional method and provides optimal treatment rules for stratified patients. (c) 2022 The Author(s). Published by Elsevier Inc.This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
查看更多>>摘要：Distributed optimization algorithms have the advantages of privacy protection and parallel computing. However, the distributed nature of these algorithms makes the system vulner-able to external attacks. This paper presents two penalty function based resilient algo-rithms for constrained distributed optimization under static and dynamic attacks. The objective function of the optimization problem is extended to nonsmooth ones and the convergence of the proposed algorithms in this case are proved under some mild condi-tions. Simulation experiments are performed and compared with some existing resilient primal-dual optimization algorithms using median-based mean estimator. For static attack, the proposed algorithm has better performance and faster convergence rate in the simulation experiments. For dynamic attack, the proposed algorithm has better perfor-mance and robustness in the simulation experiments, which illustrate that the proposed algorithms are more effective. (c) 2022 Elsevier Inc. All rights reserved.
NSTL
Elsevier