A5100352447- Bayesian Modeling and Causal Inference
- Advanced Graph Neural Networks
- Topic Modeling
- Technology Adoption and User Behaviour
- Anomaly Detection Techniques and Applications
- Bayesian Methods and Mixture Models
- Meteorological Phenomena and Simulations
- Neural Networks and Applications
- Atmospheric aerosols and clouds
- Advanced Computational Techniques and Applications
- Forecasting Techniques and Applications
- Privacy-Preserving Technologies in Data
- Multi-Criteria Decision Making
- Graph Theory and Algorithms
- Machine Learning and Data Classification
- Machine Learning and Algorithms
- Digital Marketing and Social Media
- Census and Population Estimation
- AI-based Problem Solving and Planning
- Natural Language Processing Techniques
- Precipitation Measurement and Analysis
- Decision-Making and Behavioral Economics
- Advanced Mathematical Modeling in Engineering
- Logic, Reasoning, and Knowledge
- Digital Media and Visual Art
Tianjin Agricultural University
2022-2024
Civil Aviation University of China
2023-2024
Tufts University
2019-2024
Shanghai Technical Institute of Electronics & Information
2024
Dalian University
2024
Chinese Academy of Meteorological Sciences
2009-2023
Ningxia University
2020-2023
Beijing University of Posts and Telecommunications
2023
Shanxi Normal University
2023
Daqing Oilfield General Hospital
2022
The technology acceptance model proposes that perceived ease of use and usefulness predict the information technology. Since its inception, has been tested with various applications in tens studies become most widely applied user usage. Nevertheless, reported findings on are mixed terms statistical significance, direction, magnitude. In this study, we conducted a meta-analysis based 26 selected empirical order to synthesize evidence. results suggest both correlation between acceptance,...
Transfer learning through fine-tuning a pre-trained neural network with an extremely large dataset, such as ImageNet, can significantly accelerate training while the accuracy is frequently bottlenecked by limited dataset size of new target task. To solve problem, some regularization methods, constraining outer layer weights using starting point references (SPAR), have been studied. In this paper, we propose novel regularized transfer framework DELTA, namely DEep Learning Feature Map...
Federated Learning is a new distributed learning mechanism which allows model training on large corpus of decentralized data owned by different providers, without sharing or leakage raw data. According to the characteristics dis-tribution, it could be usually classified into three categories: horizontal federated learning, vertical and transfer learning. In this paper we present solution for parallel dis-tributed logistic regression As compared with existing works, role third-party...
When formulated as an unsupervised learning problem, anomaly detection often requires a model to learn the distribution of normal data. Previous works modify Generative Adversarial Networks (GANs) by using encoder-decoders generators and apply them tasks. studies indicate that GAN ensembles are more stable than single GANs in image generation In this work, we propose construct for detection. proposed method, group interact with discriminators, so every generator gets feedback from...
Intensive field experiment is an important approach to obtain microphysical information about clouds and precipitation. From 1 July 31 August 2014, the third Tibetan Plateau Atmospheric Science Experiment was carried out comprehensive measurements of water vapor, clouds, precipitation were conducted at Naqu. The most advanced radars in China, such as Ka-band millimeter-wave cloud radar, Ku-band micro-rain C-band continuous-wave radar lidar, microwave radiometer disdrometer deployed observe...
The technology acceptance model (TAM) stipulates that both perceived ease of use (PEOU) and usefulness (PU) directly influence the end user’s behavioral intention (BI) to accept a technology. Studies have found self-efficacy is an important determinant PEOU. However, there has been no research examining relationship between BI. studies on effect PU are also rare, findings inconsistent. In this study, we incorporate Internet (ISE) into TAM as antecedent PU, PEOU, We conducted controlled...
In this study, we introduce the notion of perceived system performance (PSP) to extend technology acceptance model (TAM). We found that PSP explained 46% variation in ease use, a 50% improvement over our current understanding use while using only one predictor. also that, when is absent, TAM was validated as usual. However, present, relationship between and became insignificant, phenomenon called conditional independence . provide causal reasoning-based explanation for findings.
Federated machine learning systems have been widely used to facilitate the joint data analytics across distributed datasets owned by different parties that do not trust each others. In this paper, we proposed a novel Gradient Boosting Machines (GBM) framework SecureGBM built-up with multi-party computation model based on semi-homomorphic encryption, where every involved party can jointly obtain shared machines while protecting their own from potential privacy leakage and inferential...
The current computer-aided diagnosis for cervical cancer screening encounters issues with missing detailed information during colposcopic image segmentation and incomplete edge delineation.To overcome these challenges, this study introduces the RUC-U 2 Net architecture, which enhances through feature refinement upsampling connections.Two variants are developed: lightweight RUC + -U Net.Initially, a module that leverages an attention mechanism is proposed to improve detail capture by model's...
Graph generation is a critical task in numerous domains, including molecular design and social network analysis, due to its ability model complex relationships structured data. While most modern graph generative models utilize adjacency matrix representations, this work revisits an alternative approach that represents graphs as sequences of node set edge set. We advocate for efficient encoding propose novel representation. Based on representation, we introduce the Generative Pre-trained...
We consider the protein sequence engineering problem, which aims to find sequences with high fitness levels, starting from a given wild-type sequence. Directed evolution has been dominating paradigm in this field an iterative process generate variants and select via experimental feedback. demonstrate large language models (LLMs), despite being trained on massive texts, are secretly optimizers. With directed evolutionary method, LLM can perform through Pareto experiment-budget constrained...
Field experiment in South China was undertaken to improve understanding of cloud and precipitation properties. Measurements the vertical structures non-precipitating precipitating clouds were obtained using passive active remote sensing equipment: a Ka-band radar (CR) system, C-band frequency modulated continuous wave pointing (CVPR), microwave radiometer laser ceilometer (CEIL). CR plays key role high-level observation, whereas CVPR is important for observing low- mid-level heavy...