A5016183759- Solar Radiation and Photovoltaics
- Energy Load and Power Forecasting
- Advanced Image Fusion Techniques
- Hydrological Forecasting Using AI
- Petri Nets in System Modeling
- Tropical and Extratropical Cyclones Research
- Image Enhancement Techniques
- Data Mining Algorithms and Applications
- Plant Gene Expression Analysis
- Robotics and Automated Systems
- Robotic Path Planning Algorithms
- Cellular transport and secretion
- Natural Language Processing Techniques
- Flood Risk Assessment and Management
- Fuzzy Logic and Control Systems
- Topic Modeling
- Image and Signal Denoising Methods
- Robotics and Sensor-Based Localization
- AI-based Problem Solving and Planning
- Computational Physics and Python Applications
- Remote-Sensing Image Classification
- Remote Sensing and Land Use
- Neural Networks and Applications
- Postharvest Quality and Shelf Life Management
- Horticultural and Viticultural Research
Hohai University
2023-2025
University of Jinan
2025
East China University of Science and Technology
2024
Northwest A&F University
2023
State Key Laboratory of Resources and Environmental Information System
2023
Xi’an Jiaotong-Liverpool University
2021-2022
Shandong Agricultural University
2015-2017
Stony Brook Medicine
2016
University of Pennsylvania
2008
Hong Kong Polytechnic University
1996-2006
This paper employs the Auto-Encoding Variational Bayes (AEVB) estimator based on Stochastic Gradient (SGVB), designed to optimize recognition models for challenging posterior distributions and large-scale datasets. It has been applied mnist dataset extended form a Dynamic Bayesian Network (DBN) in context of time series. The delves into inference, variational methods, fusion Autoencoders (VAEs) techniques. Emphasis is placed reparameterization achieving efficient optimization. AEVB VAEs as...
Abstract Human selection has reshaped crop genomes. Here we report an apple genome variation map generated through sequencing of 117 diverse accessions. A comprehensive model speciation and domestication along the Silk Road is proposed based on evidence from genomic analyses. Cultivated apples likely originate Malus sieversii in Kazakhstan, followed by intensive introgressions M. sylvestris . Xinjiang China turns out to be “ancient” isolated ecotype not directly contributing domestication....
One crucial application of intelligent robotic systems is remote surveillance using a security robot. A fundamental need in the ability to automatically verify an intruder into secure or restricted area, alert personnel, and then enable them track intruder. In this article, we propose Internet-based robot system. The face recognition approach possesses "invariant" characteristics, including where facial expressions, viewing perspectives, three-dimensional poses, individual appearance,...
Ethylene and xyloglucan endotransglucosylase/hydrolase (XTH) genes were important for fruit ripening softening in 'Taishanzaoxia' apple. In this study, we found it was ACS1-1/-1 homozygotes apple, which determined the higher transcription activity of ACS1. XTH1, XTH3, XTH4, XTH5, XTH9 mainly involved early independent ethylene, while XTH2, XTH6, XTH7, XTH8, XTH10, XTH11 predominantly late dependent on ethylene. Overexpression XTH2 XTH10 tomato resulted elevated expression ethylene...
We present an automatic and integrated neural network-based tropical cyclone (TC) identification track mining system. The proposed system consists of two main modules: 1) TC pattern using oscillatory elastic graph matching model; 2) hybrid radial basis function network with time difference structural learning algorithm. For evaluation, 120 cases appeared in the period between 1985 1998 provided by National Oceanic Atmospheric Administration are being used. Comparing bureau numerical...
Apple is one of the most economically important horticultural fruit crops worldwide. It critical to gain insights into ripening and softening improve apple quality extend shelf life. In this study, forward reverse suppression subtractive hybridization libraries were generated from 'Taishanzaoxia' fruits sampled around ethylene climacteric isolate ripening- softening-related genes. A set 648 unigenes derived sequence alignment cluster assembly 918 expressed tags. According gene ontology...
Because of the stochastic and nonstationary weather changes, photovoltaic (PV) power exhibits high uncertainty fluctuation, which severely challenges operation systems. PV variations can be detected by investigating inherent properties reflections sky images. However, limited study has focused on effect image changes properties. Therefore, in this study, a novel sky-image-derived decomposition method was proposed for ultra-short-term forecasting. First, historical series were decomposed into...
Data mining is the study of how to determine underlying patterns in data help make optimal decisions on computers when database involved voluminous, hard characterize accurately and constantly changing. It deploys techniques based machine learning alongside more conventional methods. These can generate decision or prediction models actual historical data. Therefore, they represent true evidence-based support. Rainfall a good problem solve by techniques. This paper proposes an improved naive...
There is growing interest in using intelligent software agents for a variety of tasks, including navigating and retrieving information from the Internet databases, online shopping activities, user authentication, negotiation resources, decision making. We propose an integrated framework retrieval filtering context shopping. focus on applying agent technology, together with Web mining to automate series product search selection activities. It based multiagent development platform, namely,...
Weather forecasting has been one of the most challenging problems around world for more than half a century. Not only because its practical value in meteorology, but it is also typically "unbiased" time-series problem scientific researches. The paper describes methodology to short-term rainfall using neural networks. It extends previous study relying on observational data from single point station multiple sources with weather records Hong Kong region. Preprocessing procedures were important...
Internet telerobotics has emerged in recent decade with direct control and supervisory as the main teleoperation paradigms field of wheeled robot navigation. It is however difficult to apply these unknown dynamic real-world applications, while they do not provide adequate feeling interaction or a human-friendly interface human operator. This paper proposes novel interactive (active control) telecommanding, which used for Internet-based mobile teleoperation. Telecommanding involves two parts:...
The main challenge for designers of today's autonomous robotics is to build robust control programs that reliably perform complex tasks in the face real-world environmental uncertainties. This paper proposes a generalized framework behavior-based navigation strategy robots. includes goal determination, preprocessing, behavior design, arbitration, and command fusion. practical used our novel telecommanding teleoperation paradigm. experiments, including wander, advanced telecommanding, have...
Abstract Climate change-induced mega floods have become increasingly frequent worldwide. The rapid mapping and assessment of flood disasters poseurgent challenges for developing countries with poor data facilities or databases. In this study, the characteristics 2022 mega-flood in Pakistan were monitored analyzed based on multi-resources data. extent inundation throughout its impact farmlands, buildings, roads mapped using Synthetic Aperture Radar (SAR) remote sensing processing technology....