A5103056122- Machine Learning and Data Classification
- Gene expression and cancer classification
- Security in Wireless Sensor Networks
- Neural Networks and Applications
- Machine Learning in Healthcare
- Privacy-Preserving Technologies in Data
- Cloud Data Security Solutions
- Sepsis Diagnosis and Treatment
- Machine Fault Diagnosis Techniques
- Health Systems, Economic Evaluations, Quality of Life
- Anomaly Detection Techniques and Applications
- UAV Applications and Optimization
- Face and Expression Recognition
- Industrial Technology and Control Systems
- Artificial Intelligence in Healthcare
- Cryptography and Data Security
- IoT and Edge/Fog Computing
- Pharmaceutical Practices and Patient Outcomes
- Immune Response and Inflammation
- Gaze Tracking and Assistive Technology
- Bioinformatics and Genomic Networks
- Advanced Algorithms and Applications
- Face recognition and analysis
- Advanced Authentication Protocols Security
- Graph Theory and Algorithms
Shantou University
2024
Beijing University of Chemical Technology
2012-2021
Temple University
2013-2020
La Trobe University
2019
HRL Laboratories (United States)
2018
North China University of Technology
2014
Fujian Normal University
2012-2013
Abstract Background Machine learning models have been adapted in biomedical research and practice for knowledge discovery decision support. While mainstream informatics focuses on developing more accurate models, the importance of data preprocessing draws less attention. We propose Generalized Logistic (GL) algorithm that scales uniformly to an appropriate interval by a generalized logistic function fit empirical cumulative distribution data. The GL is simple yet effective; it intrinsically...
The cost of developing a new drug has increased sharply over the past years. To ensure reasonable return-on-investment, it is useful for discovery researchers in both industry and academia to identify all possible indications early pipeline molecules. For first time, we propose term computational "drug candidate positioning" or positioning", describe above process. It distinct from repositioning, which identifies uses existing drugs maximizes their value. Since many therapeutic effects are...
Abstract Background Existing feature selection methods typically do not consider prior knowledge in the form of structural relationships among features. In this study, features are structured based on into groups. The problem addressed article is how to select one representative from each group such that selected jointly discriminating classes. formulated as a binary constrained optimization and combinatorial relaxed convex-concave problem, which then transformed sequence convex problems so...
Vibration signals of rolling element bearings faults are usually immersed in background noise, which makes it difficult to detect the faults. Wavelet-based methods being used commonly can reduce some types but there is still plenty room for improvement due insufficient sparseness vibration wavelet domain. In this work, order eliminate noise and enhance weak fault detection, a new kind peak-based approach combined with multiscale decomposition envelope demodulation developed. First, preserve...
We introduce MindSpore Quantum, a pioneering hybrid quantum-classical framework with primary focus on the design and implementation of noisy intermediate-scale quantum (NISQ) algorithms. Leveraging robust support MindSpore, an advanced open-source deep learning training/inference framework, Quantum exhibits exceptional efficiency in training variational algorithms both CPU GPU platforms, delivering remarkable performance. Furthermore, this places strong emphasis enhancing operational when...
Gene expression data are widely used in classification tasks for medical diagnosis. Data scaling is recommended and helpful learning the models. In this study, we propose a algorithm to transform uniformly an appropriate interval by generalized logistic function fit empirical cumulative density of data. The proposed robust outliers, experimental results show that models learned using scaled generally outperform ones min-max mapping z-score which currently most commonly algorithms.
Software as a Service (SaaS) provides useful and convenient way for cloud users to enjoy service, but security problems have caused great influence on the development popularization of SaaS. As commercial model, SaaS involves different roles participants who could be dishonest malicious. This paper presents protocol address several issues in protect rights interests all participants. Our design is based identity-based proxy signatures from pairings. The analysis shows that presented can...
Internet of Things (IoT) is ubiquitous because its broad applications and the advance in communication technologies. The capabilities IoT also enable important role homeland security tactical missions, including Reconnaissance, Intelligence, Surveillance, Target Acquisition (RISTA). becomes most critical issue before extensive use military operations. While majority research focuses on smart devices, treatments for legacy dumb network-ready devices are lacking; moreover, deployed a hostile...
Sepsis is a serious, life-threatening condition that presents growing problem in medicine, but there still no satisfying solution for treating it. Several blood cleansing approaches recently gained attention as promising interventions target the main site of development-the blood. The focus this study an evaluation theoretical effectiveness hemoadsorption therapy and pathogen reduction therapy. This evaluated using mathematical model Murine sepsis, results over 2,200 configurations single...
Time series in healthcare practices and biomedical research are typically multivariate, i.e. multiple biomarkers observed simultaneously at a time. However, they tend to be short, noisy, unaligned, irregularly sampled, partially observed, with only limited samples. These imperfections pose challenge for mining information from data. In this work, we propose use dynamic-based representations present such imperfect multivariate time series. Specifically, an approach learn corresponding Linear...
Zero-shot learning has tremendous application value in complex computer vision tasks, e.g. image classification, localization, captioning, etc., for its capability of transferring knowledge from seen data to unseen data. Many recent proposed methods have shown that the formulation a compatibility function and generalization are crucial success zero-shot model. In this paper, we formulate softmax-based function, more importantly, propose regularized empirical risk minimization objective...
The classification model of osteoporosis was established, the diagnostic index system analyzed, and BP neural network designed. data 1261 people from 402 nuclear families were collected, including age, gender medical history. actual sampled tested by using computational function ANN, so that error could be controlled within a set range. results analysis suggested among 12 factors influencing bone mineral density, had most obvious effect, followed height, weight age. polymorphism estrogen...
It is important to select an effective subset of Acoustic Emission signal parameters achieve the real-time requirements in wireless transmits. Currently, distance wildly used abstraction. This method easily disturbed by noise and impacting resolution effect. In this paper, a new characteristic selection proposed which combined mutual information theory measurement. The approach not only considers distance, but also highlights relevance between fault classes. At same time, redundancy highly...
Abstract With the emergence and development of new automation industries such as unattended supermarkets smart picking orchards, demand for real-time systems based on embedded platforms is increasing day by day. Heterogeneous multi-core processors are widely used in modern integrated circuit design due to their advantages low power consumption high parallelism. More more implemented heterogeneous platforms. Based system, network parameters architecture jacintonet model improved, a system...
Elevated low-density lipoprotein cholesterol (LDL-C) is an influential risk factor for cardiovascular disease (CVD) morbidity/mortality. Our objective was to evaluate the impact of switches from higher-efficacy lipid-lowering therapy (HELLT) simvastatin on LDL-C levels and goal attainment among high patients in UK. This retrospective cohort study included individuals who received more than 2 months prescription following HELLT between 8/1/04 12/31/08: ezetimibe/simvastatin fixed dose...