A5101835271- Software System Performance and Reliability
- Network Security and Intrusion Detection
- Distributed Sensor Networks and Detection Algorithms
- Software Engineering Research
- Target Tracking and Data Fusion in Sensor Networks
- Anomaly Detection Techniques and Applications
- Opinion Dynamics and Social Influence
- Fault Detection and Control Systems
- Polymer-Based Agricultural Enhancements
- Translation Studies and Practices
- Sparse and Compressive Sensing Techniques
- Advanced Image Fusion Techniques
- Plant Virus Research Studies
- Advanced Image Processing Techniques
- Subtitles and Audiovisual Media
- Software Reliability and Analysis Research
- Pesticide and Herbicide Environmental Studies
- Energy Efficient Wireless Sensor Networks
- Interpreting and Communication in Healthcare
- Service-Oriented Architecture and Web Services
- Image Enhancement Techniques
- Bioinformatics and Genomic Networks
- Resilience and Mental Health
- Wireless Communication Security Techniques
- Underwater Vehicles and Communication Systems
Yulin Normal University
2024
University of Electronic Science and Technology of China
2018-2020
Microsoft Research Asia (China)
2019-2020
Chinese Academy of Sciences
2020
Southwest University
2019
Taishan University
2018
Logs are widely used by large and complex software-intensive systems for troubleshooting. There have been a lot of studies on log-based anomaly detection. To detect the anomalies, existing methods mainly construct detection model using log event data extracted from historical logs. However, we find that do not work well in practice. These close-world assumption, which assumes is stable over time set distinct events known. our empirical study shows practice, often contains previously unseen...
Online service systems have become increasingly popular. During operation of an online system, incidents (unplanned interruptions or outages the service) are inevitable. As initial step incident management, it is important to be able automatically assign report a suitable team. We call this triage, which can significantly affect efficiency and accuracy overall management. To better understand incident-triage practice in industry, we perform empirical study triage on 20 large-scale Microsoft....
In recent years, online service systems have become increasingly popular. Incidents of these could cause significant economic loss and customer dissatisfaction. Incident triage, which is the process assigning a new incident to responsible team, vitally important for quick recovery affected service. Our industry experience shows that in practice, triage not conducted only once beginning, but continuous process, engineers from different teams discuss intensively among themselves about an...
Improving plant resistance against systemic diseases remains a challenging research topic. In this study, we developed dual-action pesticide-loaded hydrogel with the capacity to significantly induce tobacco mosaic virus (TMV) infection and promote growth. We produced an alginate-lentinan-amino-oligosaccharide (ALA-hydrogel) by coating surface of alginate-lentinan drug-loaded (AL-hydrogel) amino-oligosaccharide using electrostatic action. determined formation film various approaches,...
In large-scale online service systems, incidents occur frequently due to a variety of causes, from updates software and hardware changes in operation environment. These could significantly degrade system's availability customers' satisfaction. Some are linked because they duplicate or inter-related. The can greatly help on-call engineers find mitigation solutions identify the root causes. this work, we investigate their links representative real-world incident management (IcM) system. Based...
Although tremendous efforts have been devoted to the quality assurance of online service systems, in reality, these systems still come across many incidents (i.e., unplanned interruptions and outages), which can decrease user satisfaction or cause economic loss. To better understand characteristics improve incident management process, we perform first large-scale empirical analysis collected from 18 real-world Microsoft. Surprisingly, find that although a large number could occur over short...
In the distributed detection system with multiple sensors, there are two ways for local sensors to deliver their decisions fusion center (FC): a one-bit hard decision and multiple-bit soft decision. Compared decision, has worse performance due loss of sensing information but main advantage smaller communication costs. To get tradeoff between costs performance, we propose soft–hard combination scheme clustered non-ideal channels. A is configured by fuzzy logic c-means clustering algorithm....
Hard decision (HD) and soft (SD) are two common fusion methods used in cooperative spectrum sensing (CSS). In these methods, the number of bits transmitted by each secondary user to center is always same static, namely one bit HD <tex>$\mathbf{n}(\mathbf{n}\geq 2)$</tex> SD. This paper proposes an optimal allocation scheme based on Genetic Algorithm (GA-BAS) for CSS over imperfect channels cognitive radio networks (CRNs) minimize energy consumption maximize detection probability under...
Abstract Traditional Japanese translation methods have certain disadvantages, and the introduction of artificial intelligence technology into them can enhance effect interpretation translation. In this paper, language data Twitter Facebook are used as basis to construct a corpus, GPT-2 model is constructed on Transformer for text To optimize Seq2Seq speech interpretation, Attention mechanism introduced establish model. A oral written corpus was analyze validity mentioned above. The results...
There are two main classes of decision fusion methods, namely hard (HD) and soft (SD), in which the number bits transmitted by each local sensor to centre (FC) is always same, one bit HD n (n ≥ 2) SD. However, considering that there a limit bandwidth distributed detection system, sent FC does not need be same should allocated reasonably suitably. Therefore, this study proposes an optimal allocation scheme based on memetic algorithm, could different. This aims maximise probability under for...
In this paper, a method based on non-local saturation algorithm is proposed to avoid block and halo effect for single image dehazing with dark channel prior. First we convert original from RGB color space into HSV the idea of method. Image weighted equally by size fixed window according resolution. Second utilize estimate atmospheric light value transmission rate. Then through function transmission, haze-free obtained scattering model. Comparing results existing methods, our can restore...
Hierarchical analysis for network structure can point out which communities constitute a larger group or give reasonable smaller groups within community. Numerous methods discovering community in networks divide at only one certain granularity, does not benefit hierarchical structure. clustering algorithms are the common technique that reveals multilevel analysis. In this work, we definition scores of edges according to basic idea means clustering. Based on definition, neighbors-based...