A5060333190- Energy Load and Power Forecasting
- Stock Market Forecasting Methods
- Meteorological Phenomena and Simulations
- Neural Networks and Applications
- Grey System Theory Applications
- Forecasting Techniques and Applications
- Transportation Planning and Optimization
- Statistical and numerical algorithms
- Hydrological Forecasting Using AI
- Statistical and Computational Modeling
- Pelvic and Acetabular Injuries
- Complex Systems and Time Series Analysis
- Spatial and Panel Data Analysis
- Housing Market and Economics
- Precipitation Measurement and Analysis
- Radiomics and Machine Learning in Medical Imaging
- Advanced Chemical Sensor Technologies
- Petroleum Processing and Analysis
- TiO2 Photocatalysis and Solar Cells
- Medical Imaging and Analysis
- Integrated Circuits and Semiconductor Failure Analysis
- Advanced Neural Network Applications
- Market Dynamics and Volatility
- Wind Energy Research and Development
- Electric Power System Optimization
Shanghai Changzheng Hospital
2022-2025
Central China Normal University
2012-2024
Beijing Academy of Artificial Intelligence
2024
Shanghai Artificial Intelligence Laboratory
2024
Tsinghua University
2023-2024
Xi'an Jiaotong University
2020
Hunan University
2018
Academy of Mathematics and Systems Science
2012-2014
Chinese Academy of Sciences
2012-2014
Shenyang Jianzhu University
2014
Stock e-exchange prices forecasting is an important financial problem that receiving increasing attention.This study proposes a novel three-stage nonlinear ensemble model.In the proposed model, three different types of neural-network based models, i.e.Elman network, generalized regression neural network (GRNN) and wavelet (WNN) are constructed by non-overlapping training sets further optimized improved particle swarm optimization (IPSO).Finally, neural-network-based meta-model generated...
This study aims to develop a fully automated, CT-based deep learning（DL） model segment ossified lesions of the posterior longitudinal ligament (OPLL) and measure thickness material calculate cervical spinal cord compression factor.
Data assimilation (DA) provides more accurate, physically consistent analysis fields and is used for estimating initial conditions in numerical weather forecasting. Traditional DA methods derive statistically optimal analyses model space based on Bayesian theory. However, their effectiveness limited by the difficulty of accurately background error covariances matrix B, which represents intricate interdependencies among atmospheric variables, as well standard linearity assumptions required...
Data assimilation (DA) is a statistical approach used to estimate the states of physical systems by integrating prior model predictions (background xb​) with observational data (y). This integration produces an accurate estimate, called analysis (​xa), sampling or maximizing posterior likelihood p(xxb, y). In weather forecasting, background are generated imperfect models, and p(xxb) often unknown. Observations, sourced from diverse instruments, mapped space using...
Stock e-exchange prices forecasting is an important financial problem that receiving increasing attention.This study proposes a novel three-stage nonlinear ensemble model.In the proposed model, three different types of neural-network based models, i.e.Elman network, generalized regression neural network (GRNN) and wavelet (WNN) are constructed by non-overlapping training sets further optimized improved particle swarm optimization (IPSO).Finally, neural-network-based meta-model generated...
For time series, the problem that we often encounter is how to extract patterns hidden in real world data for forecasting its future values. A single linear or nonlinear model inadequate modeling and because most of them usually contain both patterns. This study constructs a hybrid combines autoregressive integrated moving average (ARIMA) with Elman artificial neural network (ANN) short-term series. The proposed approach considers simultaneously so it can mine more precise characteristics...
The air transport industry crucially depends on traffic forecasting for supporting management decisions. In this study, a singular spectrum analysis (SSA)-based ensemble modeling approach is proposed. original passenger time series first decomposed into three components: trend, seasonal oscillations, and irregular component. trend predicted by generalized regression neural network (GRNN), whereas oscillations are radial basis function networks (RNFNs). empirical results of Hong Kong (HK)...
In recent years, the development of artificial intelligence has led to rapid advances in data-driven weather forecasting models, some which rival or even surpass traditional methods like Integrated Forecasting System (IFS) terms accuracy. However, existing models still rely on analysis fields generated by assimilation and system, hampers significance regarding both computational cost accuracy.Four-dimensional variational (4DVar) is one most popular data algorithms been adopted numerical...
Managing air passenger traffic flows is important in investing and operation of airports. However, it extremely difficult for traditional methods to analyse both the short medium terms because oscillation irregularity inherent dynamics. In this study, we design a hybrid oscillations analysis approach. The proposed method decomposes time series into different scales, making useful revealing structural breaks volatility clusters, identifying dynamic properties process at specific timescales. A...
In this article, the spatial spillover effects of nonperforming loans in commercial banks are investigated based on Chinese provinces. Panel data from 31 provinces China covering period 2005 to 2014 used study. First, we employ Moran's I statistic describe empirical evidence for presence dependencies provincial ratio (NPLR) after checking stationarity by second-generation panel unit root tests. Then construct model through a set specification tests, and calculate direct indirect explanatory...
Many efforts have been made to the development of models that able analyze and predict marine cargo volume. However, improving forecasting especially throughput time series accuracy is an important yet often difficult issue facing managers. In this study, a TEI@I methodology based hybrid model proposed. The original are decomposed different scale components using discrete wavelet technique on seasonality analysis components. All predicted by radial basis function networks due its flexible...
A high false-positive rate remains a technical glitch hindering the broad spectrum of application deep-learning-based diagnostic tools in routine radiological practice from assisting diagnosing rib fractures.To examine performance two versions software aiding radiologists fractures on chest computed tomography (CT) images.In total, 123 patients (708 fractures) were included this retrospective study. Two groups with different experience levels retrospectively reviewed images for concurrent...
Abstract For time series forecasting, the problem that we often encounter is how to increase prediction accuracy as much possible with irregular and noise data. This study proposes a novel multilayer feedforward neural network based on improved particle swarm optimization adaptive genetic operator (IPSO- MLFN). In proposed IPSO, inertia weight dynamically adjusted according feedback from particles’ best memories, acceleration coefficients are controlled by declining arccosine an increasing...
To evaluate the role of quantitative features intranodular vessels based on deep learning in distinguishing pulmonary adenocarcinoma invasiveness.This retrospective study included 512 confirmed ground-glass nodules from 474 patients with 241 precursor glandular lesions (PGL), 126 minimally invasive adenocarcinomas (MIA), and 145 (IAC). The blood were reconstructed noncontrast computed tomography images using learning-based region-segmentation region-growing techniques. presence was evaluated...
It is very significant for us to predict future energy consumption accurately. As China's annual time series, the sample size relatively small. This study combines traditional auto-regressive model with group method of data handling (GMDH) suitable small prediction, and proposes a novel GMDH based (GAR) model. can finish modeling process in self-organized manner, including finding optimal complexity model, determining order estimating parameters. Further, four different GAR models, AS-GAR,...
Motivation: Thyroid-associated ophthalmopathy (TAO) is characterized by accumulation of collagen in extraocular muscle. CEST-MRI can evaluate the content focusing on amide compound. However, CEST effect small and sensitive to low image SNR. A vendor-provided deep learning reconstruction (DLR) algorithm dramatically increase Goal(s): Investigate if distinguish inactive from active TAO impact DLR its diagnostic performance. Approach: 11 12 were enrolled. imaging was reconstructed with...