2024 |
Han X, Chu Y, Wang K, Wang L, Yue L, Ding W, 'TAILOR: InTer-feAture distinctIon fiLter fusiOn pRuning', Information Sciences, 665 (2024) [C1]
Filter pruning is an effective method for reducing the size of convolutional neural networks without sacrificing performance. Most filter pruning methods prioritize filters with h... [more]
Filter pruning is an effective method for reducing the size of convolutional neural networks without sacrificing performance. Most filter pruning methods prioritize filters with high information content, but fail to consider that filters with low information content might capture essential features. Moreover, we have discovered that the distinctions among feature maps generated by filters can identify crucial features. Based on this insight, we propose a novel pruning method called inTer-feAture distinctIon fiLter fusiOn pRuning (TAILOR), which fuses the feature distinctions between filters. TAILOR randomly selects multiple filter sets within a convolutional layer and calculates the output feature maps of the next convolutional layer of these sets. Subsequently, an intelligent distinction optimization scheme is proposed to obtain the optimal filter set for filter pruning, which supplants the original convolutional layer. Experimental results indicated that the inter-feature distinctions among filters significantly affect filter pruning. TAILOR outperforms state-of-the-art filter-pruning methods in terms of model prediction accuracy, floating-point operations, and parameter scale. For instance, with VGG-16, TAILOR achieves a 73.89% FLOPs reduction by removing 91.85% of the parameters, while improving accuracy by 0.36% on the CIFAR-10.
|
|
Nova |
2024 |
Xu Y, Li T, Yang Y, Chen W, Yue L, 'An adaptive category-aware recommender based on dual knowledge graphs', Information Processing and Management, 61 (2024) [C1]
Combining the knowledge graph (KG) with the personalized item recommendation has become an important method to improve user experience. In the personalized item recommendation, us... [more]
Combining the knowledge graph (KG) with the personalized item recommendation has become an important method to improve user experience. In the personalized item recommendation, users have their preferences on categories that influence their choices of items. In order to fully use category information, we explicitly focus on their impact on user preference and run through the whole recommendation process. We construct two dual knowledge graphs (KG-UI and -UC). Based on them, we propose KG-CICEF, a recommendation system based on knowledge graph aggregation and user preference modeling. Our model effectively captures user preferences for explored and unexplored item categories by aggregating information from two types of knowledge graphs. We convert user preference over unexplored item categories to the cross-item-category exploration factor (CEF). We utilize CEF to build a category-wise loss function for the item recommendation. For consistency, we also propose a category-based negative sampling mechanism to optimize this loss function. Experimental results on three benchmark datasets demonstrate that KG-CICEF achieves significant improvements over the state-of-the-art methods, and the case study validates the effectiveness of CEF in item recommendations.
|
|
Nova |
2023 |
Guo L, Wang L, Han X, Yue L, Zhang Y, Gao M, 'ROCM: A Rolling Iteration Clustering Model Via Extracting Data Features', Neural Processing Letters, 55 3899-3922 (2023) [C1]
The allocation of boundary points and low-density clusters has become an essential part of clustering research. Most of the recent improved methods that focused on identifing allo... [more]
The allocation of boundary points and low-density clusters has become an essential part of clustering research. Most of the recent improved methods that focused on identifing allocation of points have not addressed the issue of specific data point assignment in terms of the data¿s distribution feature. In this article, a rolling iteration clustering model (ROCM) was proposed for assigning the specific data point by extracting the feature of data points. In this model, data points were transformed into multiple units with different distribution structures, and then each unit¿s dispersion used to discover representative groups was analyzed. Sparse data were clustered based on the proposed self-expansion principle to effectively capture boundary points and assign points at joint. Furthermore, the rolling iteration module avoided the over-partitioning and chaining effect and discovered clusters with diverse shapes and densities. Experimental results of twenty-two datasets proved the effectiveness of the proposed method. ROCM has better performance than other state-of-the-art methods.
|
|
|
2023 |
Li B, Yue L, Tao C, Han X, Calvanese D, Amagasa T, 'Preface', Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 13422 LNCS v-vi (2023) |
|
|
2023 |
Amagasa T, Calvanese D, Han X, Li B, Tao C, Yue L, 'Preface', Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 13423 LNCS v-vi (2023) |
|
|
2023 |
Amagasa T, Calvanese D, Han X, Li B, Tao C, Yue L, 'Preface', Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 13421 LNCS v-vi (2023) |
|
|
2023 |
Shi Z, Wang S, Yue L, Zhang Y, Adhikari BK, Xue S, et al., 'Dual-core mutual learning between scoring systems and clinical features for ICU mortality prediction', Information Sciences, 637 (2023) [C1]
Perpetually improving mortality prediction in intensive care units (ICUs) via the implementation of eHealth evaluation approaches has become a major research hotspot in the field ... [more]
Perpetually improving mortality prediction in intensive care units (ICUs) via the implementation of eHealth evaluation approaches has become a major research hotspot in the field of medical data mining for the purpose of saving lives. Recently, researchers have attempted to achieve improved prediction accuracy by using only deep learning-based techniques. However, some problems remain. (1) Most of the existing methods utilize independent clinical features to predict mortality by eliminating the correlations between the latent features, but this technique may fail to comprehensively capture and evaluate the statuses of patients. (2) Several clinical features are needed to ensure strong prediction accuracy, but most methods only use static features that are not extendable. (3) An effective practical framework that unifies traditional ICU scoring systems and state-of-the-art deep learning methods to predict mortality is also lacking. (4) Moreover, the interpretability of existing deep learning-based methods needs to be further improved. Therefore, we propose a novel dual-core mutual learning framework (DMLF) between ICU scoring systems and clinical features for mortality prediction. In particular, we mutually utilize sequential organ failure assessment (SOFA) scores and clinical measurement features to learn a unified model for enhancing the accuracy and interpretability of our DMLF. Experiments conducted on five real-world disease datasets show that the DMLF achieves significantly better prediction accuracy and area under the receiver operating characteristic curve (AUROC) values than six baselines and four state-of-the-art methods. Moreover, clinicians utilize a familiarized SOFA system to conduct mortality prediction and achieve increased interpretability, which benefits the adoption of the proposed framework in real clinical scenarios.
|
|
Nova |
2023 |
Chen W, Zhang WE, Yue L, 'Death comes but why: A multi-task memory-fused prediction for accurate and explainable illness severity in ICUs', WORLD WIDE WEB-INTERNET AND WEB INFORMATION SYSTEMS, 26 4025-4045 (2023) [C1]
|
|
Nova |
2022 |
Peng T, Han R, Cui H, Yue L, Han J, Liu L, 'Distantly Supervised Relation Extraction using Global Hierarchy Embeddings and Local Probability Constraints', KNOWLEDGE-BASED SYSTEMS, 235 (2022) [C1]
|
|
|
2022 |
Wang YD, Chen WT, Pi DC, Yue L, 'Adaptive Multi-Hop Reading on Memory Neural Network with Selective Coverage Mechanism for Medication Recommendation', Tien Tzu Hsueh Pao/Acta Electronica Sinica, 50 943-953 (2022)
Medication recommendation aims to make effective prescriptions based on electronic healthcare records (EHRs) of patients, and assists caregivers in clinical decision making. Obtai... [more]
Medication recommendation aims to make effective prescriptions based on electronic healthcare records (EHRs) of patients, and assists caregivers in clinical decision making. Obtaining temporal patterns of patient conditions as well as contextual information contained in EHRs are the key issues for the success of recommendation. Existing methods do not take the difference in the amount of medical records of different patients into account, and fails to change the focus or number of iterations during information extraction according to personalized patient conditions. To address these problems, the medication recommendation model adaptive multi-hop reading with selective coverage mechanism (AMHSC) is proposed. The model stores encoded temporal patterns with memory neural networks (MemNN), and applies the selective coverage mechanism to balance attention weights over selected information during the attentive multi-hop reading on MemNN. Meanwhile, AMHSC adaptively determines the number of reading hops on MemNN according to personalized patient conditions. Experiments on real-world clinical dataset demonstrate that AMHSC successfully derives important information from EHRs to build informative patient representations for medication recommendation.
|
|
|
2022 |
Zhang Y, Peng T, Han R, Han J, Yue L, Liu L, 'Synchronously tracking entities and relations in a syntax-aware parallel architecture for aspect-opinion pair extraction', APPLIED INTELLIGENCE, 52 15210-15225 (2022) [C1]
|
|
|
2021 |
Wang Y, Chen W, Pi D, Yue L, 'Adversarially regularized medication recommendation model with multi-hop memory network', KNOWLEDGE AND INFORMATION SYSTEMS, 63 125-142 (2021) [C1]
|
|
|
2021 |
Shi Z, Wang S, Yue L, Pang L, Zuo X, Zuo W, Li X, 'Deep dynamic imputation of clinical time series for mortality prediction', INFORMATION SCIENCES, 579 607-622 (2021) [C1]
|
|
|
2021 |
Yue L, Shen H, Wang S, Boots R, Long G, Chen W, Zhao X, 'Exploring BCI Control in Smart Environments: Intention Recognition Via EEG Representation Enhancement Learning', ACM TRANSACTIONS ON KNOWLEDGE DISCOVERY FROM DATA, 15 (2021) [C1]
|
|
|
2020 |
Yijia Z, Zhenkun S, Wanli Z, Lin Y, Shining L, Xue L, 'Joint Personalized Markov Chains with social network emb e dding for cold -start recommendation', NEUROCOMPUTING, 386 208-220 (2020) [C1]
|
|
|
2020 |
Yue L, Tian D, Chen W, Han X, Yin M, 'Deep learning for heterogeneous medical data analysis', WORLD WIDE WEB-INTERNET AND WEB INFORMATION SYSTEMS, 23 2715-2737 (2020) [C1]
|
|
|
2020 |
Shi Z, Zuo W, Liang S, Zuo X, Yue L, Li X, 'IDDSAM: An Integrated Disease Diagnosis and Severity Assessment Model for Intensive Care Units', IEEE ACCESS, 8 15423-15435 (2020) [C1]
|
|
|
2019 |
Yue L, Chen W, Li X, Zuo W, Yin M, 'A survey of sentiment analysis in social media', KNOWLEDGE AND INFORMATION SYSTEMS, 60 617-663 (2019) [C1]
|
|
|
2018 |
Yue L, Sun X-X, Gao W-Z, Feng G-Z, Zhang B-Z, 'Multiple Auxiliary Information Based Deep Model for Collaborative Filtering', JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY, 33 668-681 (2018) [C1]
|
|
|
2018 |
Zhao X, Li J, Wang R, He F, Yue L, Yin M, 'General and Species-Specific Lysine Acetylation Site Prediction Using a Bi-Modal Deep Architecture', IEEE ACCESS, 6 63560-63569 (2018) [C1]
|
|
|
2017 |
Yue L, Shi Z, Han J, Wang S, Chen W, Zuo W, 'Multi-factors based sentence ordering for cross-document fusion from multimodal content', NEUROCOMPUTING, 253 6-14 (2017) [C1]
|
|
|
2017 |
Yang Y, Xu Y, Han J, Wang E, Chen W, Yue L, 'Efficient traffic congestion estimation using multiple spatio-temporal properties', NEUROCOMPUTING, 267 344-353 (2017) [C1]
|
|
|
2015 |
Yue L, Zuo W, Peng T, Wang Y, Han X, 'A fuzzy document clustering approach based on domain-specified ontology', DATA & KNOWLEDGE ENGINEERING, 100 148-166 (2015)
|
|
|
2015 |
Guo L, Zuo W, Peng T, Yue L, 'Text Matching and Categorization: Mining Implicit Semantic Knowledge from Tree-Shape Structures', Mathematical Problems in Engineering, 2015 (2015)
The diversities of large-scale semistructured data make the extraction of implicit semantic information have enormous difficulties. This paper proposes an automatic and unsupervis... [more]
The diversities of large-scale semistructured data make the extraction of implicit semantic information have enormous difficulties. This paper proposes an automatic and unsupervised method of text categorization, in which tree-shape structures are used to represent semantic knowledge and to explore implicit information by mining hidden structures without cumbersome lexical analysis. Mining implicit frequent structures in trees can discover both direct and indirect semantic relations, which largely enhances the accuracy of matching and classifying texts. The experimental results show that the proposed algorithm remarkably reduces the time and effort spent in training and classifying, which outperforms established competitors in correctness and effectiveness.
|
|
|
2014 |
Yue L, Zuo W, Feng L, Guo L, 'OMFM: A Framework of Object Merging Based on Fuzzy Multisets', MATHEMATICAL PROBLEMS IN ENGINEERING, 2014 (2014)
|
|
|
2012 |
Yang X, Zou C, Yue L, Gao R, 'Research on food complains document classification based-on topic', Journal of Software, 7 1687-1693 (2012)
In this paper, we design a classifier based-on topic for food complain documents, and take a series of measures to the implementation process. In order to accomplish feature reduc... [more]
In this paper, we design a classifier based-on topic for food complain documents, and take a series of measures to the implementation process. In order to accomplish feature reduction, the filter method named term filtering for independent topic features is proposed to compress each topic feature vector. We introduce the created food ontology as background knowledge and to expand the semantic of complaint documents with the aid of HowNet. So we can supplement effective information and improve the effect of text classification. In addition, we take account of different importance between title and body in the text, considering that title can stand out the topic of text better than the textual body. Consequently, we separately calculate the word frequency which words are in textual title and body. The experiments show that it is necessary to consider the different importance between textual title and body, and we can achieve good feature reduction effect using the proposed filter method, and the classification performance get obvious improvement after the process of term expanding. © 2012 ACADEMY PUBLISHER.
|
|
|
2012 |
Yang XQ, Yue L, Ma ZQ, Lv YH, Martinez FS, 'An ontology-driven approach based on fuzzy equivalence relation for document clustering analysis', ICIC Express Letters, 6 1321-1327 (2012)
Document clustering is an effective way to detect similar clusters of documents in nature. The objective of this paper is to develop an effective methodology to cluster food compl... [more]
Document clustering is an effective way to detect similar clusters of documents in nature. The objective of this paper is to develop an effective methodology to cluster food complaint documents with the guidance of ontology and fuzzy set theory. Here, the terms in the controlled vocabulary of our food ontology describe the hazards in dairy product and will be used to guide feature selection. Then, the fuzzy compatibility relation will be constructed in k-dimensional semantic space and a fuzzy equivalence relation based on the fuzzy compatibility relation can be obtained. Finally, a suitable ¿-cut value is selected to determine the best number of clusters. Several documents have been tested and the results show our methodology is effective and accurate.
|
|
|
2011 |
Yang X, Yue L, Liu C, Zou C, Tian Y, 'A knowledge-driven approach for document clustering', ICIC Express Letters, Part B: Applications, 2 1123-1129 (2011)
Concurrent with progress in food supervision, an overwhelming number of textual complaints are accumulating in web pages. The representation of these documents by a set of domain-... [more]
Concurrent with progress in food supervision, an overwhelming number of textual complaints are accumulating in web pages. The representation of these documents by a set of domain-specific terminologies and their relationships is the result of this process. The objective of this paper is to develop an effective method to cluster food complaint documents based on a user-specified ontology. First, we generalize some concepts with a controlled vocabulary in our food ontology based on Hownet to extract the document features. Especially, the food ontology will be extended during the procedure of feature extraction. Term clustering techniques and latent semantic analysis (LSA) are used for document clustering. The experimental results show the effectiveness of the methodology. © 2011 ISSN 2185-2766.
|
|
|