A5101676462- Hydrology and Drought Analysis
- Climate variability and models
- Topic Modeling
- Climate change impacts on agriculture
- Bayesian Modeling and Causal Inference
- Data Management and Algorithms
- Computational and Text Analysis Methods
- Video Analysis and Summarization
- Domain Adaptation and Few-Shot Learning
- UAV Applications and Optimization
- Natural Language Processing Techniques
- Geochemistry and Geologic Mapping
- Rough Sets and Fuzzy Logic
- Advanced Image and Video Retrieval Techniques
- Hate Speech and Cyberbullying Detection
- Reinforcement Learning in Robotics
- Multimodal Machine Learning Applications
- Flood Risk Assessment and Management
Nanjing University of Information Science and Technology
2022-2025
Chinese Academy of Surveying and Mapping
2014-2016
Coordinating multiple agents with diverse tasks and changing goals without interference is a challenge. Multi-Agent Reinforcement Learning (MARL) aims to develop effective communication joint policies using group learning. Some of the previous approaches required each agent maintain set networks independently, resulting in no consideration interactions. Joint work causes receiving information unrelated their own tasks. Currently, different task divisions are often grouped by action tendency,...
Abstract. Networks play the role of a high-level language, as is seen in Artificial Intelligence and statistics, because networks are used to build complex model from simple components. These years, Bayesian Networks, one probabilistic networks, powerful data mining technique for handling uncertainty domains. In this paper, we apply Augmented Naive Bayes (BAN) texture classification High-resolution satellite images put up new method construct network topology structure terms training...
Abstract. Image classification will still be a long way in the future, although it has gone almost half century. In fact, researchers have gained many fruits image domain, but there is distance between theory and practice. However, some new methods artificial intelligence domain absorbed into draw on strength of each to offset weakness other, which open up prospect. Usually, networks play role high-level language, as seen Artificial Intelligence statistics, because are used build complex...
<abstract> <p>Most current deep learning-based news headline generation models only target domain-specific data. When a new domain appears, it is usually costly to obtain large amount of data with reference truth on the for model training, so text trained by traditional supervised approaches often do not generalize well domain—inspired idea transfer learning, this paper designs cross-domain method based distribution alignment, intermediate redistribution, and zero-shot learning...
Online shopping has become a crucial way to encourage daily consumption, where the User-generated, or crowdsourced product comments, can offer broad range of feedback on e-commerce products. As result, integrating critical opinions major attitudes from comments provide valuable for marketing strategy adjustment product-quality monitoring. Unfortunately, scarcity annotated ground truth integrated comment, limited gold integration reference, incurred infeasibility regular...