A5100603879- Higher Education and Teaching Methods
- Blind Source Separation Techniques
- Recommender Systems and Techniques
- Advanced Computational Techniques and Applications
- Educational Technology and Assessment
- Topic Modeling
- Education and Work Dynamics
- Reinforcement Learning in Robotics
- Educational Technology and Pedagogy
- Bayesian Modeling and Causal Inference
- Multi-Agent Systems and Negotiation
- Neural Networks and Applications
- Data Management and Algorithms
- Rough Sets and Fuzzy Logic
- Spectroscopy and Chemometric Analyses
- Data Mining Algorithms and Applications
- Service-Oriented Architecture and Web Services
- Imbalanced Data Classification Techniques
- Human Mobility and Location-Based Analysis
- Digital Media and Visual Art
- Fault Detection and Control Systems
- Machine Learning in Healthcare
- Industrial Technology and Control Systems
- Speech and Audio Processing
- Traffic Prediction and Management Techniques
Guangdong University of Technology
2014-2024
Zhejiang University
2010-2024
Wuhan University of Technology
2007-2024
Midea Group (China)
2023
Soochow University
2017-2021
Shanghai Maritime University
2018
Xuzhou Central Hospital
2017
Agricultural Information Institute
2013-2017
Chinese Academy of Agricultural Sciences
2013-2017
Beijing Institute of Technology
2012-2013
Decision tree (and its extensions such as Gradient Boosting Trees and Random Forest) is a widely used machine learning algorithm, due to practical effectiveness model interpretability. With the emergence of big data, there an increasing need parallelize training process decision tree. However, most existing attempts along this line suffer from high communication costs. In paper, we propose new called \emph{Parallel Voting Tree (PV-Tree)}, tackle challenge. After partitioning data onto number...
Domain adaptation on time series data is an important but challenging task. Most of the existing works in this area are based learning domain-invariant representation with help restrictions like MMD. However, such extraction a non-trivial task for data, due to complex dependence among timestamps. In detail, fully dependent series, small change lags or offsets may lead difficulty domain invariant extraction. Fortunately, stability causality inspired us explore structure data. To reduce...
In this paper, a new method based on deep learning for robotics autonomous navigation is presented. Different from the most traditional methods fixed models, convolutional neural network (CNN) modelling technique in Deep selected to extract feature inspired by working pattern of biological brain. This model has muti-layer features where ambient scenes can be recognized and useful information such as location door identified. The extracted used robot navigation, so does approach target...
Causal discovery from observational data is a fundamental problem in science. Though the linear non-Gaussian acyclic model (LiNGAM) has shown promising results various applications, it still faces following challenges with multiple latent confounders: 1) how to detect confounders and 2) uncover causal relations among observed variables. To address these two challenges, we propose hybrid method for LiNGAM (MLCLiNGAM). First, utilize constraint-based learn skeleton. Second, identify...
With the boom in online clothing e-commerce, various web portals and mobile applications apply recommendation methods to improve sales consumer satisfaction based on massive historical records big data era. This study examined collaborative filtering algorithms embedded typical for clothing. The test dataset is constructed with a real-world large-scale instance from one of largest business-to-consumer e-commerce platforms (www.taobao.com) China. Considering purchasing times inverse user...
This paper focuses on a specific stochastic shortest path (SSP) problem, namely the punctuality problem. It aims to determine that maximizes probability of arriving at destination before specified deadline. The popular solution this problem always formulates it as cardinality minimization by considering its data-driven nature, which is approximately solved 1 , -norm relaxation. To address accurately, we consider special character in cardinality-based and reformulate introducing additional...
Large Language Models (LLMs) tend to prioritize adherence user prompts over providing veracious responses, leading the sycophancy issue. When challenged by users, LLMs admit mistakes and provide inaccurate responses even if they initially provided correct answer. Recent works propose employ supervised fine-tuning (SFT) mitigate issue, while it typically leads degeneration of LLMs' general capability. To address challenge, we a novel pinpoint tuning (SPT), where region-of-interest modules are...
Social network systems such as FaceBook and YouTube have played a significant role in capturing both explicit implicit user preferences for different items the form of ratings tags. This forms quaternary relationship among users, items, tags ratings. Existing utilized only ternary relationships users-items-ratings, or users-items-tags to derive their recommendations. In this paper, we show that are insufficient provide accurate Instead, model 4-order tensor cast recommendation problem...
This paper delineates the visual speech recognition (VSR) system introduced by NPU-ASLP-LiAuto (Team 237) in first Chinese Continuous Visual Speech Recognition Challenge (CNVSRC) 2023, engaging fixed and open tracks of Single-Speaker VSR Task, track Multi-Speaker Task. In terms data processing, we leverage lip motion extractor from baseline1 to produce multi-scale video data. Besides, various augmentation techniques are applied during training, encompassing speed perturbation, random...
Modern automatic speech recognition (ASR) model is required to accurately transcribe diverse signals (from different domains, languages, accents, etc) given the specific contextual information in various application scenarios. Classic end-to-end models fused with extra language perform well, but mainly data matching scenarios and are gradually approaching a bottleneck. In this work, we introduce Seed-ASR, large (LLM) based model. Seed-ASR developed on framework of audio conditioned LLM...
With numerous news available on the Web every day, needs for effective ranking algorithms to satisfy users' requirements are continuously increasing. Though it has attracted lots of interests in commercial world, little academic research been done about ranking. In this paper, we introduce a virtual graph model describe properties news. Based model, propose algorithm which can fully exploit reinforcement between sources, topics and articles. Our be processed line. experiments, shows good...
The recommendation system, relying on historical observational data to model the complex relationships among users and items, has achieved great success in real-world applications. Selection bias is one of most important issues existing data-based approaches, which actually caused by multiple types unobserved exposure strategies (e.g., promotions holiday effects). Though various methods have been proposed address this problem, they are mainly implicit debiasing techniques but not explicitly...