A5071709543- Advanced Graph Neural Networks
- Complex Network Analysis Techniques
- Functional Brain Connectivity Studies
- Face and Expression Recognition
- Topic Modeling
- Tensor decomposition and applications
- stochastic dynamics and bifurcation
- Probabilistic and Robust Engineering Design
- Advanced Clustering Algorithms Research
- Advanced Neuroimaging Techniques and Applications
- Text and Document Classification Technologies
- Video Surveillance and Tracking Methods
- Recommender Systems and Techniques
- Bioinformatics and Genomic Networks
- Chaos control and synchronization
- Advanced Image and Video Retrieval Techniques
- Multimodal Machine Learning Applications
- Domain Adaptation and Few-Shot Learning
- Machine Learning in Healthcare
- EEG and Brain-Computer Interfaces
- Traffic Prediction and Management Techniques
- Human Mobility and Location-Based Analysis
- Privacy-Preserving Technologies in Data
- Sentiment Analysis and Opinion Mining
- Natural Language Processing Techniques
Lehigh University
2019-2025
Chongqing University of Posts and Telecommunications
2010-2025
Yuntianhua Group (China)
2022-2025
University of Electronic Science and Technology of China
2024
Anhui University of Technology
2022-2024
Kunming University of Science and Technology
2023
Tangshan People's Hospital
2022
Tianjin Medical University
2022
Beihang University
2021
University of Pennsylvania
2019-2021
Text classification is the most fundamental and essential task in natural language processing. The last decade has seen a surge of research this area due to unprecedented success deep learning. Numerous methods, datasets, evaluation metrics have been proposed literature, raising need for comprehensive updated survey. This paper fills gap by reviewing state-of-the-art approaches from 1961 2021, focusing on models traditional We create taxonomy text according involved used feature extraction...
Deep neural networks (DNNs) have been widely applied to various applications, including image classification, text generation, audio recognition, and graph data analysis. However, recent studies shown that DNNs are vulnerable adversarial attacks. Though there several works about attack defense strategies on domains such as images natural language processing, it is still difficult directly transfer the learned knowledge due its representation structure. Given importance of analysis, an...
Multi-view clustering can explore common semantics from multiple views and has attracted increasing attention. However, existing works punish objectives in the same feature space, where they ignore conflict between learning consistent reconstructing inconsistent view-private information. In this paper, we propose a new framework of multi-level for contrastive multi-view to address aforementioned issue. Our method learns different levels features raw features, including low-level high-level...
CNNs, RNNs, GCNs, and CapsNets have shown significant insights in representation learning are widely used various text mining tasks such as large-scale multi-label classification. Most existing deep models for classification consider either the non-consecutive long-distance semantics or sequential semantics. However, how to coherently take them into account is still far from studied. In addition, most methods treat output labels independent medoids, ignoring hierarchical relationships among...
Generative commonsense reasoning which aims to empower machines generate sentences with the capacity of over a set concepts is critical bottleneck for text generation. Even state-of-the-art pre-trained language generation models struggle at this task and often produce implausible anomalous sentences. One reason that they rarely consider incorporating knowledge graph can provide rich relational information among concepts. To promote ability generation, we propose novel augmented model...
Graph representation learning has attracted increasing research attention. However, most existing studies fuse all structural features and node attributes to provide an overarching view of graphs, neglecting finer substructures' semantics, suffering from interpretation enigmas. This paper presents a novel hierarchical subgraph-level selection embedding-based graph neural network for classification, namely SUGAR, learn more discriminative subgraph representations respond in explanatory way....
Multi-view clustering is an important research topic due to its capability utilize complementary information from multiple views. However, there are few methods consider the negative impact caused by certain views with unclear structures, resulting in poor multi-view performance. To address this drawback, we propose <u>s</u>elf-supervised discriminative feature learning for <u>d</u>eep <u>m</u>ulti-<u>v</u>iew <u>c</u>lustering (SDMVC). Concretely, deep autoencoders applied learn embedded...
Pretrained Foundation Models (PFMs) are regarded as the foundation for various downstream tasks with different data modalities. A PFM (e.g., BERT, ChatGPT, and GPT-4) is trained on large-scale which provides a reasonable parameter initialization wide range of applications. BERT learns bidirectional encoder representations from Transformers, large datasets contextual language models. Similarly, generative pretrained transformer (GPT) method employs Transformers feature extractor using an...
Abstract Drug development is time‐consuming and expensive. Repurposing existing drugs for new therapies an attractive solution that accelerates drug at reduced experimental costs, specifically Coronavirus Disease 2019 (COVID‐19), infectious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2). However, comprehensively obtaining productively integrating available knowledge big biomedical data to effectively advance deep learning models still challenging repurposing...
Mapping the connectome of human brain using structural or functional connectivity has become one most pervasive paradigms for neuroimaging analysis. Recently, Graph Neural Networks (GNNs) motivated from geometric deep learning have attracted broad interest due to their established power modeling complex networked data. Despite superior performance in many fields, there not yet been a systematic study how design effective GNNs network To bridge this gap, we present BrainGB, benchmark analysis...
Cluster analysis plays an indispensable role in machine learning and data mining. Learning a good representation is crucial for clustering algorithms. Recently, deep (DC), which can learn clustering-friendly representations using neural networks (DNNs), has been broadly applied wide range of tasks. Existing surveys DC mainly focus on the single-view fields network architectures, ignoring complex application scenarios clustering. To address this issue, article, we provide comprehensive survey...
There has been growing interest in developing more effective learning machines for tensor classification. At present, most of the existing machines, such as support machine (STM), involve nonconvex optimization problems and need to resort iterative techniques. Obviously, it is very time-consuming may suffer from local minima. In order overcome these two shortcomings, this paper, we present a novel linear higher-order (SHTM) which integrates merits C-support vector (C-SVM) rank-one...
Text classification is the most fundamental and essential task in natural language processing. The last decade has seen a surge of research this area due to unprecedented success deep learning. Numerous methods, datasets, evaluation metrics have been proposed literature, raising need for comprehensive updated survey. This paper fills gap by reviewing state-of-the-art approaches from 1961 2021, focusing on models traditional We create taxonomy text according involved used feature extraction...
In this paper, we propose an online multi-view clustering algorithm, OMVC, which deals with large-scale incomplete views. We model the problem as a joint weighted NMF and process data chunk by to reduce memory requirement. OMVC learns latent feature matrices for all views pushes them towards consensus. further increase robustness of learned in via lasso regularization. To minimize influence incompleteness, dynamic weight setting is introduced give lower weights incoming missing instances...
GNNs have been widely used in deep learning on graphs. They learn effective node representations. However, most methods ignore the heterogeneity. Methods designed for heterogeneous graphs, other hand, fail to complex semantic representations because they only use meta-paths instead of meta-graphs. Furthermore, cannot fully capture content-based correlations, as either do not self-attention mechanism or it consider immediate neighbors each node, ignoring higher-order neighbors. We propose a...
Events are happening in real world and time, which can be planned organized for occasions, such as social gatherings, festival celebrations, influential meetings, or sports activities. Social media platforms generate a lot of real-time text information regarding public events with different topics. However, mining is challenging because typically exhibit heterogeneous texture metadata often ambiguous. In this article, we first design novel event-based meta-schema to characterize the semantic...
Multi-view clustering, a long-standing and important research problem, focuses on mining complementary information from diverse views. However, existing works often fuse multiple views' representations or handle clustering in common feature space, which may result their entanglement especially for visual representations. To address this issue, we present novel VAE-based multi-view framework (Multi-VAE) by learning disentangled Concretely, define view-common variable view-peculiar variables...
Understanding the interconnected relationships of large-scale information networks like social, scholar and Internet Things is vital for tasks recommendation fraud detection. The vast majority real-world are inherently heterogeneous dynamic, containing many different types nodes edges can change drastically over time. dynamicity heterogeneity make it extremely challenging to reason about network structure. Unfortunately, existing approaches inadequate in modeling real-life dynamical as they...
Depression is one of the most common mental illnesses, and symptoms shown by patients are different, making it difficult to diagnose in process clinical practice pathological research. Although researchers hope that artificial intelligence can contribute diagnosis treatment depression, traditional centralized machine learning methods need aggregate patient data, data privacy with illness needs be strictly confidential, which hinders algorithms' application. To solve problem medical this...