A5100386190- Topic Modeling
- Machine Learning in Healthcare
- Advanced Database Systems and Queries
- Web Data Mining and Analysis
- Semantic Web and Ontologies
- Digital Media and Visual Art
- Sentiment Analysis and Opinion Mining
- Educational Technology and Pedagogy
- Environmental remediation with nanomaterials
- Biomedical Text Mining and Ontologies
- Advanced oxidation water treatment
- Data Management and Algorithms
- Electrostatic Discharge in Electronics
- Advanced Malware Detection Techniques
- Software Testing and Debugging Techniques
- Image and Signal Denoising Methods
- Artificial Intelligence in Healthcare
- Context-Aware Activity Recognition Systems
- Advanced Image Processing Techniques
- Consumer Market Behavior and Pricing
- Auction Theory and Applications
- Forensic Toxicology and Drug Analysis
- Metabolomics and Mass Spectrometry Studies
- Cloud Computing and Resource Management
- Blind Source Separation Techniques
Beijing University of Posts and Telecommunications
2024
Beijing Normal University
2024
Hubei University of Medicine
2024
Tokyo Institute of Technology
2024
National University of Defense Technology
2009-2024
Zhejiang Lab
2024
Nankai University
2011-2023
Wuxi Taihu Hospital
2023
Data Assurance and Communication Security
2021-2022
Wuhan Textile University
2022
The missing values, appear in most of multivariate time series, prevent advanced analysis series data. Existing imputation approaches try to deal with values by deletion, statistical imputation, machine learning based and generative imputation. However, these methods are either incapable dealing temporal information or multi-stage. This paper proposes an end-to-end model E²GAN impute series. With the help discriminative loss squared error loss, can incomplete nearest generated complete at...
This paper proposes a new method for computing page importance, referred to as BrowseRank. The conventional approach compute importance is exploit the link graph of web and build model based on that graph. For instance, PageRank such an algorithm, which employs discrete-time Markov process model. Unfortunately, might be incomplete inaccurate with respect data determining because links can easily added deleted by content creators. In this paper, we propose using 'user browsing graph' created...
IFTTT is a popular trigger-action programming platform whose applets can automate more than 400 services of IoT devices and web applications. We conduct an empirical study using combined approach analyzing data collected for 6 months performing controlled experiments custom testbed. profile the interactions among different entities, measure how are used by end users, test performance applet execution. Overall we observe fast growth ecosystem its increasing usage automating IoT-related tasks,...
Photovoltaic (PV) is receiving increasing attention due to its sustainability and low carbon footprint. However, the penetration level of PV still relatively because intermittency. This uncertainty can be handled by accurate forecasting, which requires high-quality solar data. Nevertheless, up 40% data found missing, significantly worsens quality letter proposes a novel solarGAN method for multivariate imputation, in necessary modifications are made on input generative adversarial network...
A facile, general, and efficient one-pot approach to obtaining narrow or monodisperse, highly cross-linked, surface-functionalized, "living" polymer microspheres with uniformly cross-linked structures by atom transfer radical precipitation polymerization (ATRPP) is described for the first time. The simple introduction of (ATRP) mechanism into system allows direct generation their number-average diameters ranging from 0.73 3.25 μm polydispersity indices being typically lower than 1.01....
Embeddings of medical concepts such as medication, procedure and diagnosis codes in Electronic Medical Records (EMRs) are central to healthcare analytics. Previous work on concept embedding takes EMRs words documents respectively. Nevertheless, models miss out the temporal nature EMR data. On one hand, two consecutive do not indicate they temporally close, but correlations between them can be revealed by time gap. other scopes often vary greatly (e.g., common cold diabetes). In this paper,...
The prevailing use of both images and text to express opinions on the web leads need for multimodal sentiment recognition. Some commonly used social media data containing short few images, such as tweets product reviews, have been well studied. However, it is still challenging predict readers' after reading online news articles, since articles often more complicated structures, e.g., longer images. To address this problem, we propose a layout-driven attention network (LD-MAN) recognize in an...
This paper is concerned with the joint allocation of bid price and campaign budget in sponsored search. In this application, an advertiser can create a number campaigns set for each them. campaign, he/she further several ad groups keywords prices. Data analysis shows that many advertisers are dealing very large campaigns, keywords, prices at same time, which poses great challenge to optimality their management. As result, budgets some might be too low achieve desired performance goals while...
Cloud gaming has been very popular recently, but providing satisfactory experiences to players at a modest cost is still challenging. Colocating several games onto one server could improve utilization. To enable efficient colocations while Quality of Service (QoS) guarantees, precise quantification performance interference among colocated required. However, achieving such prediction challenging for due the complexity introduced by contention on many shared resources across CPU and GPU....
Digital capacity is of vital importance for the digital natives to live in both physical and virtual world. However, seldom research on learning could be found literature. This paper aims investigate framework status Chinese students' according a proposed framework. A Learning Capacity Questionnaire (DLCQ) was developed capacity. Based survey results analysis data, 4-factor 10-subscale questionnaire validated. The relationship each subscales discussed students reported.
Finding treatments that accelerate peripheral nerve regeneration, prolongation, and functional recovery remains a challenging task. Platelet-rich plasma (PRP) contains numerous growth factors active proteins, low-dose ultrashort waves (USWs) stimulate the formation of nerve-nourishing vessels, which are powerful for regeneration. The goal this study was to evaluate synergistic effects serial ultrasound-guided PRP injections combined with USWs radiation on regeneration in crush injury model....
Automated medical named entity recognition and normalization are fundamental for constructing knowledge graphs building QA systems. When it comes to text, the annotation demands a foundation of expertise professionalism. Existing methods utilize active learning reduce costs in corpus annotation, as well multi-task strategy model correlations between different tasks. However, existing models do not take task-specific features tasks diversity query samples into account. To address these...
General practitioners are supposed to be better diagnostics detect patients with serious diseases earlier, and conduct early interventions appropriate referrals of patients. However, in the current general practice, primary lack sufficient clinical experiences, correct rate disease diagnosis is low. To assist diagnosis, this paper proposes a multi-label hierarchical classification method based on graph neural network, which integrates medical knowledge electronic health record (EHR) data...
Embeddings of medical concepts such as medication, procedure and diagnosis codes in Electronic Medical Records (EMRs) are central to healthcare analytics. Previous work on concept embedding takes EMRs words documents respectively. Nevertheless, models miss out the temporal nature EMR data. On one hand, two consecutive do not indicate they temporally close, but correlations between them can be revealed by time gap. other scopes often vary greatly (e.g., \textit{common cold}...
Abstract Flunitrazepam (FNZ) is a potent hypnotic, sedative, and amnestic drug used to treat severe insomnia. In our recent study, FNZ metabolic profiles were investigated carefully. Six authentic human urine samples purified using solid phase extraction (SPE) without enzymatic hydrolysis, extracts then analyzed by liquid chromatography‐Q exactive‐HF hybrid quadrupole‐Orbitrap‐mass spectrometry (LC‐QE‐HF‐MS), the full scan positive ion mode targeted MS/MS (ddms2) technique make accurate mass...