A5023025722- Data Mining Algorithms and Applications
- Rough Sets and Fuzzy Logic
- Data Management and Algorithms
- Advanced Computational Techniques and Applications
- Recommender Systems and Techniques
- Heat Transfer Mechanisms
- Turbomachinery Performance and Optimization
- Text and Document Classification Technologies
- Advanced Clustering Algorithms Research
- Fluid Dynamics and Turbulent Flows
- Advanced Text Analysis Techniques
- Web Data Mining and Analysis
- Advanced Database Systems and Queries
- Imbalanced Data Classification Techniques
- Metallurgy and Material Forming
- Music and Audio Processing
- Aluminum Alloy Microstructure Properties
- Speech Recognition and Synthesis
- Higher Education and Teaching Methods
- Topic Modeling
- Anomaly Detection Techniques and Applications
- Natural Language Processing Techniques
- Hydraulic flow and structures
- Complex Network Analysis Techniques
- Advanced Steganography and Watermarking Techniques
North China University of Technology
2016-2025
Jiangnan University
2010-2025
Shanghai Maritime University
2025
Nantong University
2022-2024
Beijing Information Science & Technology University
2005-2024
Aero Engine Corporation of China (China)
2020-2024
Northeastern University
2024
State Grid Corporation of China (China)
2016-2024
University of Science and Technology of China
2014-2024
China University of Mining and Technology
2010-2024
Deep learning model (primarily convolutional networks and LSTM) for time series classification has been studied broadly by the community with wide applications in different domains like healthcare, finance, industrial engineering IoT. Meanwhile, Transformer Networks recently achieved frontier performance on various natural language processing computer vision tasks. In this work, we explored a simple extension of current gating, named Gated (GTN) multivariate problem. With gating that merges...
Mining high utility itemsets (HUI) is an interesting research problem in the field of data mining and knowledge discovery. Recently, bio-inspired computing has attracted considerable attention, leading to development new algorithms for HUIs. These have shown good performance terms efficiency, but are not guaranteed find all HUIs a database. That is, quality comparatively poor number discovered To solve this problem, framework based on bioinspired proposed. This approach adjusts standard...
To address the challenges faced by coal miners when encountering collapsed pillars in local hard rock formations, we researched ultra-deep hole loosening blasting technology. overcome issues related to inapplicability of existing zoning theories and lack a solid foundation for borehole design sites, introduced model theoretical calculation formula based on thick-walled cylinder theory. Building this foundation, proposed coupled uncoupled charge methods that take into account strain rate...
Mining high utility itemsets is one of the most important research issues in data mining owing to its ability consider nonbinary frequency values items transactions and different profit for each item. Although a number relevant approaches have been proposed recent years, they incur problem producing large candidate itemsets. In this paper, authors propose an efficient algorithm, namely BAHUI (Bitmap-based Algorithm High Utility Itemsets), with bitmap database representation. BAHUI, used...
Due to its anonymity, there has been a dramatic growth of underground drug markets hosted in the darknet (e.g., Dream Market and Valhalla). To combat trafficking (a.k.a. illicit trading) cyberspace, is an urgent need for automatic analysis participants markets. However, one key challenges that traffickers (i.e., vendors) may maintain multiple accounts across different or within same market. address this issue, paper, we propose develop intelligent system named uStyle-uID leveraging both...
High utility itemset mining (HUIM) is the task of finding all items set, purchased together, that generate a high profit in transaction database. In past, several algorithms have been developed to mine itemsets (HUIs). However, most them cannot properly handle exponential search space while HUIs when size database and total number increases. Recently, evolutionary heuristic were designed HUIs, which provided considerable performance improvement. they can still long runtime some may miss many...
The aim of sequential pattern mining (SPM) is to discover potentially useful information from a given sequence. Although various SPM methods have been investigated, most these focus on all the patterns. However, users sometimes want mine patterns with same specific prefix pattern, called co-occurrence pattern. Since rule can make better use results SPM, and obtain recommendation performance, this paper addresses issue maximal nonoverlapping (MCoR) proposes MCoR-Miner algorithm. To improve...
High-impedance fault detection poses significant challenges for distribution network maintenance and operation. We propose a dual-path neural high-impedance detection. To enhance feature extraction, we use Gramian Angular Field algorithm to transform 1D zero-sequence voltage signals into 2D images. Our dual-branch simultaneously processes both representations: the CNN extracts spatial features from transformed images, while GRU captures temporal raw signals. optimize model performance,...
We assessed whether constructing a mathematical knowledge graph for question-answering system or course recommendation system, Named Entity Recognition (NER), is indispensable. The accuracy of its recognition directly affects the actual performance these subsequent tasks. In order to improve entity and provide effective support functionalities, this paper adopts latest pre-trained language model, LERT, combined with Bidirectional Gated Recurrent Unit (BiGRU), Iterated Dilated Convolutional...
Background: Distant supervision employs external knowledge bases to automatically match with text, allowing for the automatic annotation of sentences. Although this method effectively tackles challenge manual labeling, it inevitably introduces noisy labels. Traditional approaches typically employ sentence-level attention mechanisms, assigning lower weights sentences mitigate their impact. But approach overlooks critical importance information flow between Additionally, previous treated an...
Single-image super-resolution (SISR) methods based on convolutional neural networks (CNNs) have achieved breakthrough progress in reconstruction quality. However, their high computational costs and model complexity limited applications resource-constrained devices. To address this, we propose the MSWSR (multi-scale wavelet super-resolution) method, a lightweight multi-scale feature selection network that exploits both symmetric asymmetric patterns. achieves efficient extraction fusion...
ABSTRACT Learning path recommendation is crucial for guiding learners through a series of courses in logical sequence based on their previous learning experiences. This particularly important improving outcomes massive open online (MOOCs) diverse learners. Because both the historical and recommended paths can be represented as sequential patterns (SPs); it reasonable to approach this problem SP mining (SPM). In addition support, we incorporate three factors, that is, course days, grades...