查看更多>>摘要：The Linear time-invariant dynamic models are widely adopted in the industry. In the machine learning domain, such models are known as time-invariant continuous-state hidden Gauss-Markov models. Their super-class, the linear time-varying dynamic models, have relatively sparse applications as predictive models and classifiers of time series. This is typically due to the model complexity and the need for a significantly larger training set than time-invariant models. Without a large training set, a better modeling performance is counteracted by a less robust model. In this paper, we propose the continuity preference of the time-varying parameters of the model, which significantly reduces the required amount of training data while maintaining the modeling performance. We also derive a simple modification of the Expectation-Maximization algorithm incorporating continuity in parameters. The modified algorithm shows robust learning performance. The model performance is demonstrated by experiments on real 6-axis robotic manipulators in a laboratory, the Skoda Auto car producer body shop, and also on a public benchmark data set. (C) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：The recommender system is the key approach to alleviate the data explosion problem. Recently, with the rapid development of deep learning, there are several researches of employing deep neural networks (DNNs) on recommender systems. Most of these methods tend to capture the complex mapping relations between user-item representation and matching score via DNNs. These methods are mainly a pyramid structure which maps relations into low-dimensional space and then predicts the result by logistic regression. However, partial relations may be linearly indivisible in low-dimensional space. As we know, data that are hard to be separated in low-dimensional space can become much easier after being mapped into a high-dimensional space. Hence, motivated by the ladder network, we propose a Symmetric Deep Neural Networks (SDNN) with lateral connections, which can learn relations in both high-dimensional and low-dimensional spaces simultaneously. Moreover, considering that deep neural network is very inefficient in catching low-rank relations between users and items, we further combine SDNN with an improved deep matrix factorization model into a unified framework, and name this new model DualCF. Extensive experiments on three benchmark datasets are conducted and the results verify the effectiveness of SDNN and DualCF over state-of-the-art models for implicit feedback prediction. (C) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：Three-way decision (3WD) theory has been widely used in solving multi-attribute decision-making (MADM) problems, however the calculation method for semantic interpretations of loss functions and conditional probabilities has not been reasonably addressed. Thus, by virtue of existing studies on 3WD and MADM, a novel 3WD method to MADM is explored in the current paper, which is titled the 3W-MADM method. First, a pair of novel pre-order relations are designed by virtue of the classic TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) method, and a new scheme for calculating conditional probabilities is further established. Then, the calculation scheme of loss functions and thresholds is put forward via attribute evaluation values and ideal solutions. In particular, several primary steps and the algorithm of the 3W-MADM method are summed up. Finally, the practicality and validity of the proposed method are demonstrated by a realistic case of lymphography from the UCI database. Moreover, the stability and superiority of the presented method are shown by several experimental analyses. (C) 2022 Elsevier Inc. All rights reserved.
Zhang, JinglanAtukorale, Ajantha S.Subhashini, L. D. C. S.Li, Yuefeng...
21页
查看更多>>摘要：A three-way decision framework based on fuzzy concepts can better understand uncertain decision boundaries by automatically dividing opinions into three decision regions: positive, negative, and boundary regions. However, when the fuzzy values of the opinions are very close, the boundary regions tend to be large. The difficult problem is how to significantly reduce the boundary regions while maintaining high classification accuracy. The most useful opinion features include fuzzy concepts and semantic features (or patterns). Unlike fuzzy concepts, semantic features are represented by using semantic patterns that frequently appear in opinions. We also observe a low statistical correlation between semantic patterns and fuzzy concepts. Therefore, this paper proposes a novel opinion classification method that integrates semantic patterns with fuzzy concepts in a three-way decision framework. The new method can increase the discriminative power of classifying opinions using features. Experimental results verify that the integration of fuzzy concepts and semantic patterns has better classification performance than using fuzzy concepts alone. This also shows that fusing multiple features is an effective solution to represent opinions more meaningfully, thereby effectively improving classification accuracy and reducing uncertain boundaries in three-way opinion classification. (C) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：The presence of missing data is a challenging issue in processing real-world datasets. It is necessary to improve the data quality by imputing the missing values so that effective learning from data can be achieved. Recently, deep learning has become the most powerful type of machine learning techniques, which can be used for discovering the hidden knowledge that exists in a large dataset to make accurate predictions. In this paper, we propose an imputation method that involves using a convolutional neural network to impute the missing values. The missing value of each instance is imputed essentially by using a trained kernel. The weights of the kernel are determined by learning from the given data that are arranged spatially in the data matrix. The kernel carries out a weighted sum of neighboring elements in an array for imputing the missing values. In addition, in the absence of the true values with which the missing values are expected to be replaced, a loss function is designed without the need to know the true value. Our method is evaluated on UCI datasets in comparison with state-of-the-art methods. The experimental results show that the proposed approach performs closely to or better than other methods. (C) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：One of the challenges of learning algorithms is data imbalance, especially data intrinsic dif-ficulties such as borderline, rare, and outlier examples. In order to improve the perfor-mance of learning algorithms in such cases, various solutions have been proposed, one of which is ensemble learning. Despite the good results of ensemble learning, their results are not interpretable. In some domains, such as medicine, banking, and telecommunica-tions, the interpretability of the final model results is very important. For this purpose, rule-based learning algorithms and decision trees are two appropriate approaches. Therefore, an interPretable Ensemble Rule Learning (PERL) model is proposed in this paper. PERL has three main phases: sampling, rule learning, and rule processing. We reduce the negative effects of data intrinsic difficulties in each phase by an appropriate solution. PERL is compared with several interpretable ensemble models on ten imbalanced real -world datasets. According to the obtained results, PERL outperforms significantly the other models based on the f1-score measure. (c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：In this article, we study aggregation operators on shadowed sets. In particular, since shadowed sets can be obtained as approximations of fuzzy sets, we explore the relationships between aggregation operators on fuzzy sets and corresponding operators on shadowed sets. We focus on studying conditions under which the approximations of fuzzy sets into shadowed sets represent an homomorphism with respect to the corresponding aggregation operators, and we propose classes of fuzzy set operators that correspond to (and satisfy the same properties as) specific shadowed set operators.(c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：Nowadays, tremendous information technology industries resort to cloud servers to store data with an outsourcing approach to extend their storage and computation power. This, however, also leads to privacy and security issues of unprotected data against curious cloud servers. The most common solution currently is to encrypt the data before uploading it. The problem is that encrypted data is out of control and users also cannot transfer the searching job to cloud serves anymore. Specifically, the order preserving encryption (OPE) provides an efficient solution to the order of plaintexts. Existing OPE schemes focus on single-dimensional data, and fail to effectively process multi-dimensional data. In this paper, we propose a multi-dimensional data order preserving encryption scheme MDOPE allowing fine-grained multi-dimensional range queries. Our scheme constructs query indexes for each dimension of the data based on the order preserving encryption network. In particular, the proposed scheme ensures that no external entity, including the cloud server, can obtain additional information other than the order of ciphertexts during the whole query process.(C) 2022 Elsevier Inc. All rights reserved.
Macababayao, Ivan Cedric H.Cabarle, Francis George C.de la Cruz, Ren Tristan A.Zeng, Xiangxiang...
20页
查看更多>>摘要：Spiking Neural P (SN P) systems are membrane computing systems that are abstracted from the behavior of spiking neurons, or brain cells. These systems take advantage of various features, such as the ability of neurons to forget, the ability of neurons to create and remove synapses, and many others. Some variants of SN P systems are (1) SN P systems with Structural Plasticity, which include the ability to create and delete synapses, and (2) SN P systems with Rules on Synapses, which associates rules with synapses instead of with neurons. The main results of this work show that for SN P systems, having only one type of regular expression in the entire system is sufficient for universality. Moreover, for the two variants of SN P systems mentioned above, having a maximum of one rule per neuron and one regular expression in the system is sufficient for universality. For normal forms with such parameters, e.g. number of rules per neuron, types of regular expressions in the system, our universality results are optimal. We also show some optimisations on the types of neurons in a system, involving the removal of some unbounded neurons in favour of simpler and bounded neurons.(c) 2022 Published by Elsevier Inc.
NSTL
Elsevier