A5100680717- Face and Expression Recognition
- Cloud Computing and Resource Management
- Tensor decomposition and applications
- Advanced Image and Video Retrieval Techniques
- Computational Physics and Python Applications
- Distributed and Parallel Computing Systems
- Image Retrieval and Classification Techniques
- Sparse and Compressive Sensing Techniques
- Spine and Intervertebral Disc Pathology
- Spinal Fractures and Fixation Techniques
- Parallel Computing and Optimization Techniques
- Water Systems and Optimization
- Video Surveillance and Tracking Methods
- Software-Defined Networks and 5G
- Medical Imaging and Analysis
- Text and Document Classification Technologies
- Advanced Radiotherapy Techniques
- Advanced Neural Network Applications
- Digital Radiography and Breast Imaging
- Superconducting Materials and Applications
- Water Treatment and Disinfection
- Pelvic and Acetabular Injuries
- Infrared Target Detection Methodologies
- Radiomics and Machine Learning in Medical Imaging
- Machine Learning and Data Classification
Peking University
2021-2024
Peking University Third Hospital
2024
South China University of Technology
2021-2024
Peking University Cancer Hospital
2021
University of Warwick
2018-2021
Guangdong University of Technology
2018-2020
Dalian Neusoft University of Information
2020
Tsinghua University
2018-2019
Wuhan Institute of Technology
2016
Xiangtan University
2016
Short-time water demand forecasting is essential for optimal control in a distribution system (WDS). Current methods (e.g., time-series models and conventional artificial neural networks) have limited power practice due to the nonlinear nature of changes demand. In particular, 15-min time-step may not be accurate when using models. To tackle this problem, paper investigates potential deep learning short-term forecasting, developing gated recurrent unit network (GRUN) model forecast 15 min 24...
Multi-view clustering exploits the complementary information of different views for comprehensive data analysis. Recently, graph learning techniques with low-dimensional embedding have been developed to learn consensus affinity multi-view clustering. However, projecting into space has often resulted in compression information, which is insufficient learning. To address this challenge, paper proposes a Collaborative Embedding Learning via Tensor (CELT) method, learns intra-view graphs each...
Unsupervised feature selection is a vital yet challenging topic for effective data learning. Recently, 2-D methods show good performance on image analysis by utilizing the structure information of image. Current usually adopt sparse regularization to spotlight key features. However, such scheme introduces additional hyperparameter needed pruning, limiting applicability unsupervised algorithms. To overcome these challenges, we design filter estimate weight features selection. Theoretical...
Unsupervised feature selection is vital in explanatory learning and remains challenging due to the difficulty of formulating a learnable model. Recently, graph embedding has gained widespread popularity unsupervised learning, which extracts low-dimensional representation based on structure. Nevertheless, such an scheme for will distort original features spatial transformation by extraction. To address this problem, paper proposes collaborative model via jointly using soft-threshold learning....
Current cross-modal retrieval methods heavily rely on accurate semantic labels or sample similarity measurements, and need to search for the nearest samples among all in huge space, severely limiting application stratifying large-scale high-dimensional multimodal data. To tackle with issues, this paper proposes an unsupervised method bypass semanticwise supervision samplewise from a standpoint of featurewise matching, named by dual hashing coding (UDC). It jointly learns codes tagging...
Traditional clustering methods rely on pairwise affinity to divide samples into different subgroups. However, high-dimensional small-sample (HDLSS) data are affected by the concentration effects, rendering traditional metrics unable accurately describe relationships between samples, leading suboptimal results. This article advances proposition of employing high-order affinities characterize multiple sample as a strategic means circumnavigate effects. We establish nexus order constructing...
Abstract In this paper, we investigate the influential factors that impact on performance when tasks are co-running a multicore computers. Further, propose machine learning-based prediction framework to predict of tasks. particular, two frameworks developed for types task in our model: repetitive (i.e., arrive at system repetitively) and new submitted first time). The difference between which is have historical running information while do not prior knowledge about Given limited tasks, an...
This study was to propose and validate an efficient streamlined quality assurance (QA) method with a single phantom setup check performances of patient positioning guidance systems including six-degree-of-freedom (6DoF) couch, X-ray modalities (kV-kV, MV-MV CBCT), optical surface imaging system (AlignRT), lasers distance indicator (ODI).The QA based on pseudo-patient treatment plan using the AlignRT cube phantom. The first randomly set up initial position offsets were acquired by CBCT....
Hashing technology has exhibited great cross-modal retrieval potential due to its appealing efficiency and storage effectiveness. Most current supervised methods heavily rely on accurate semantic supervision, which is intractable for annotations with ever-growing sample sizes. By comparison, the existing unsupervised similarity preservation strategies intensive computational costs compensate lack of guidance, causes these lose power bridge gap. Furthermore, both kinds approaches need search...
To confirm the effect of surgery on spinal column biomechanics and to provide theoretical support for advantages disadvantages different surgical methods their clinical efficacy.
Due to the inherent high-dimensional characteristics of genomic data, traditional single metric/kernel-based clustering methods fail accurately perform data analysis. To address this issue, we propose a multi-kernel with tensor fusion on Grassmann manifold (MKCTM). Specifically, multiple kernel functions are employed map into different spaces and utilize representations capture their high-order relationships. By introducing low-rank constraint, maximize correlation among kernels while...
The All-Pairs Shortest Paths (APSP) is a fundamental graph problem aiming to find the shortest path between any two nodes in graph. In this paper, new method presented solve APSP for big graphs on distributed systems. method, partitioned judiciously and then processed parallel. particular, first pre-processed prepare partition computation stages. After into smaller sub-graphs, traditional algorithm, such as Floyd-Warshall algorithm or Dijkstra's can be used each sub-graph. Finally, through...