A5100364503- Topic Modeling
- Multi-Criteria Decision Making
- Machine Learning and Data Classification
- Advanced Multi-Objective Optimization Algorithms
- Advanced Bandit Algorithms Research
- Stock Market Forecasting Methods
- Geochemistry and Geologic Mapping
- Machine Learning and Algorithms
- Personal Information Management and User Behavior
- Recommender Systems and Techniques
- Natural Language Processing Techniques
- Probabilistic and Robust Engineering Design
- Business Process Modeling and Analysis
- Healthcare Operations and Scheduling Optimization
- Fuzzy Systems and Optimization
- Sentiment Analysis and Opinion Mining
- Optimization and Mathematical Programming
- Advanced Text Analysis Techniques
- Online Learning and Analytics
- Machine Learning and ELM
- Domain Adaptation and Few-Shot Learning
- Quantum Information and Cryptography
- Forecasting Techniques and Applications
- Statistical Methods and Bayesian Inference
- Open Source Software Innovations
Chongqing Normal University
2024-2025
Xidian University
2025
École Polytechnique Fédérale de Lausanne
2024
Shaanxi University of Science and Technology
2023
Jingdong (China)
2021-2023
Nanjing University
2016-2022
Chengdu University of Technology
2022
Northwestern Polytechnical University
2020-2022
Nanjing University of Science and Technology
2017-2021
East China Normal University
2017-2021
Xinyu Hua, Mitko Nikolov, Nikhil Badugu, Lu Wang. Proceedings of the 2019 Conference North American Chapter Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers). 2019.
Matrix concentration inequalities have attracted much attention in diverse applications such as linear algebra, statistical estimation, combinatorial optimization, etc. In this paper, we present new Bernstein depending only on the first moments of random matrices, whereas previous are heavily relevant to and second moments. Based those results, analyze empirical risk minimization presence label noise. We find that many popular losses used can be decomposed into two parts, where part won't...
Abstract Peer‐to‐peer energy trading enhances distribution network resilience by reducing demand from central power plants and enabling distributed resources to support critical loads after extreme events. However, adequate reserves main grids are still required ensure real‐time balance in networks due the uncertainty renewable generation. This paper introduces a novel two‐stage joint reserve market for prosumers, wherein local flexible fully utilized manage generation uncertainty. In...
Abstract Quantum communication in the terahertz (THz) band is an important technology for next‐generation high‐capacity wireless networks, and inter‐satellite quantum key distribution low‐earth‐orbit (LEO) one of critical research directions. A continuous‐variable measurement‐device‐independent (CV‐MDI‐QKD) with photon subtraction secure link communications, derive security bounds protocol, proposed performance asymptotic limit, addition to obtaining a tighter protocol bound by considering...
Summary In some survival analysis of medical studies, there are often long‐term survivors who can be considered as permanently cured. The goals in these studies to estimate the noncured probability whole population and hazard rate susceptible subpopulation. When covariates present happens practice, understand covariate effects on is equal importance. existing methods limited parametric semiparametric models. We propose a two‐component mixture cure model with nonparametric forms for both...
Problem definition: Collaboration is important in services but may lead to interruptions. Professionals exercise discretion on when preempt individual tasks switch collaborative tasks. Academic/practical relevance: Discretionary task switching can introduce changeover times resuming the preempted and, thus, increase total processing time. Methodology: We analyze and quantify how collaboration, through interruptions discretionary changeovers, affects an episodal workflow model that captures...
Summary Combining multiple studies is frequently undertaken in biomedical research to increase sample sizes for statistical power improvement. We consider the marginal model regression analysis of repeated measurements collected several similar with potentially different variances and correlation structures. It great importance examine whether there exist common parameters across study-specific models so that simpler models, sensible interpretations, meaningful efficiency gain can be...
In web-based scenarios, new users and items frequently join the recommendation system over time without prior events. addition, always hold dynamic diversified preferences. Therefore, cold-start diversity are two serious challenges of system. Recent works show that these problems can be effectively solved by contextual multi-armed bandit (CMAB) algorithms which consider coldstart process as a game. But existing methods only treat either or arms, causing lower accuracy on other side. this...
Click-through rate (CTR) prediction plays a crucial role in sponsored search advertising (search ads). User click behavior usually showcases strong comparison patterns among relevant/competing items within the user awareness. Explicit awareness could be characterized by sequence modeling, which however suffers from issues such as cold start, noise and hidden channels. Instead, this paper, we study problem of modeling implicit about items. We notice that candidate CTR model play surrogates...
This paper focuses on the scientific problem of quantitative mineralization prediction at large depth in Zaozigou gold deposit, west Qinling, China. Five geological and geochemical indicators are used to establish model. Machine learning Deep algorithms employed for 3D Mineral Prospectivity Mapping (MPM). Especially, Student Teacher Ore-induced Anomaly Detection (STOAD) model is proposed based knowledge distillation (KD) idea combined with Auto-encoder (DAE) network Compared DAE, STOAD uses...
Psychiatric practice routinely uses semistructured and/or unstructured free text to record the behavior and mental state of patients. Many these data are unstructured, lack standardization, difficult use for analysis. Thus, it is quantitatively analyze a patient's illness trajectory over time his or her responsiveness treatment, also compare different patients quantitatively. In this article, experts in field psychiatry, along with machine learning models, have collaboratively transformed...
In this technology world, education is also becoming one of the basic necessities human life like food, shelter, and clothes. Even in day-to-day daily activities, world moving toward an automated process using developments. Some developments activities are smartphone, internet home office appliances. To cope with these advanced technologies, persons must have educational qualification to understand operate those appliances easily. Apart from this, helps person develop their personal growth...
The emergence of large language models (LLMs) has substantially influenced natural processing, demonstrating exceptional results across various tasks. In this study, we employ ``Introspective Tips" to facilitate LLMs in self-optimizing their decision-making. By introspectively examining trajectories, LLM refines its policy by generating succinct and valuable tips. Our method enhances the agent's performance both few-shot zero-shot learning situations considering three essential scenarios:...
After entering the 21st century, China’s national economy has shown a rapid growth momentum, comprehensive transportation system been continuously improved, road traffic infrastructure made remarkable achievements, and modern logistics industry also risen rapidly grown rapidly, which greatly changed market demand for transport hubs. The hub is main node of network, passenger freight distribution transport, organizational center interconnection other modes development transport. Highway an...