A5100617917- Plant Water Relations and Carbon Dynamics
- Gene expression and cancer classification
- Hydrology and Drought Analysis
- Advanced Harmonic Analysis Research
- Molecular Biology Techniques and Applications
- Bioinformatics and Genomic Networks
- Environmental and Agricultural Sciences
- Nonlinear Partial Differential Equations
- Climate variability and models
- Data Mining Algorithms and Applications
- Advanced Causal Inference Techniques
- Hydrology and Watershed Management Studies
- Advanced Computational Techniques and Applications
- Tree-ring climate responses
- Research studies in Vietnam
- Advanced Graph Neural Networks
- Indoor and Outdoor Localization Technologies
- Mathematical Analysis and Transform Methods
- Complex Network Analysis Techniques
- Explainable Artificial Intelligence (XAI)
- Radio Wave Propagation Studies
- Topic Modeling
- Education and Learning Interventions
- Machine Learning in Healthcare
- Natural Language Processing Techniques
Jilin University
2019-2024
Sichuan Agricultural University
2023
Xi'an University of Technology
2023
Zhejiang Normal University
2022
Guangdong University of Foreign Studies
2022
Jilin International Studies University
2021
Institute of Electronics
2021
Baidu (China)
2020
University of Sheffield
2020
Jilin Medical University
2020
Graph structured data has wide applicability in various domains such as physics, chemistry, biology, computer vision, and social networks, to name a few. Recently, graph neural networks (GNN) were shown be successful effectively representing because of their good performance generalization ability. However, explaining the effectiveness GNN models is challenging task complex nonlinear transformations made over iterations. In this paper, we propose GraphLIME, local interpretable model...
With the rapid development of social media services in recent years, relational data are explosively growing. The signed network, which consists a mixture positive and negative links, is an effective way to represent friendly hostile relations among nodes, can users or items. Because features associated with node network usually incomplete, noisy, unlabeled, high-dimensional, feature selection important procedure eliminate irrelevant features. However, existing network-based methods linear...
Evapotranspiration is an important process in the hydrologic cycle. In this study, different types of daily evapotranspiration were accurately estimated and comparatively analyzed for 2020–2021 gully area Loess Plateau, China with multisource observation data. The actual was directly observed by eddy covariance system then compared simulated HYDRUS-1D, they showed reasonable agreement similar variability characteristics it a correlation coefficient 0.71. FAO-56 Penman-Monteith equation used...
In this paper, a hierarchical attention network is proposed to generate utterance-level embeddings (H-vectors) for speaker identification and verification. Since different parts of an utterance may have contributions identities, the use structure aims learn related information locally globally. approach, frame-level encoder are applied on segments input individual segment vectors. Then, level vectors construct representation. To evaluate effectiveness data NIST SRE2008 Part1 used training,...
Language learning in classrooms has been studied from different perspectives.Among numerous focuses, the present study aims to examine three dimensions-classroom activities, classroom anxiety and teacher roles as well their interactions by investigating 49 students of an English class at university.Quantitative instruments 5-likert scales were used measure number frequency activities class, level perceptions roles.Results showed that was "often" dominated interactive coupled "sometimes" with...
Estimating individual treatment effect (ITE) from observational data has attracted great interest in recent years, which plays a crucial role decision-making across many high-impact domains such as economics, medicine, and e-commerce. Most existing studies of ITE estimation assume that different units at play are independent do not influence each other. However, social science experiments have shown there often exist levels interactions between data, especially networked environment. As...
Based on the risk characters and conception of compensating regulation, identification model was established. Monte-Carlo random simulation for multifactor evaluation built applied to analyses regulation cascade reservoirs main stream Yellow River in combination with indexes, 19 integrated schemes 2010 water level year evaluated. The results indicate that analysis developed study is reasonable feasible.
Cancer staging, grading and subtyping all represent important problems for precision diagnosis, treatment mechanistic studies of cancer. The majority the existing computational methods solve this problem via multi-classification differential gene-expressions cancer samples specific classes (Stages, Grades subtypes) vs. controls. However, performance such classification techniques is generally not satisfactory since discerning power expression patterns in classifications limited. We present...
Tea pests and diseases are among the primary constraints in tea industry. However, practitioners often rely on books internet for information, leading to fragmented time-consuming searches. Constructing a question-answering system based knowledge graph of can address these issues. This study utilizes deep learning model BERT-BiLSTM-CRF automatically extract triplets, enabling automatic construction automated it. research facilitates rapid development agricultural sector provides solutions...
Abstract It is significant but challenging to explore a subset of robust biomarkers distinguish cancer from normal samples on high-dimensional imbalanced biological omics data. Although many feature selection methods addressing high dimensionality and class imbalance have been proposed, they rarely pay attention the fact that most classes will dominate final decision-making when dataset imbalanced, leading instability it expands downstream tasks. Because causality invariance, causal...
In this paper, we investigate the $L^{2}$ boundedness of Fourier integral operator $T_{\phi,a}$ with rough symbol $a \in L^{\infty} S^{m}_{\rho}$ and phase $\phi \Phi^{2}$ which satisfies $\big| \{ x: |\nabla_{\xi} \phi(x,\xi) - y| \leq r \} \big| C(r^{n-1}+r^{n})$ for any $\xi,y \mathbb{R}^{n}$ $r \gt 0$. We obtain that is bounded on $L^2$ if $m \lt \rho(n-1)/2 n/2$ when $0 \rho 1/2$ or -(n+1)/4$ $1/2 1$. When $\rho = 0$ $n 1$, condition $m$ sharp. Moreover, maximal wave a special class...
Tea production involves several stages, usually pests and diseases can negatively impact the quality of tea reduce harvest, limiting industry's development.Therefore, it is important to unify knowledge on diseases.Unfortunately, current graph construction for relies mainly semi-automated manual methods, resulting in inefficiency failing meet demands.This research combines three model Bidirectional Encoder Representation from Transformers, Bi-directional Long Short-Term Memory, Conditional...
Let $\mathcal{L}$ be a Schr\"odinger operator of the form $\mathcal{L} = -\Delta+V$ acting on $L^2(\mathbb R^n)$ where nonnegative potential $V$ belongs to reverse H\"older class $B_q$ for some $q\geq n.$ $L^{p,\lambda}(\mathbb{R}^{n})$, $0\le \lambda<n$ denote Morrey space $\mathbb{R}^{n}$. In this paper, we will show that function $f\in L^{2,\lambda}(\mathbb{R}^{n})$ is trace solution ${\mathbb L}u=u_{t}+{\mathcal{L}}u=0, u(x,0)= f(x),$ $u$ satisfies Carleson-type condition...