Low Earth Orbit (LEO) satellites can be used to assist maritime wireless communications for data transmission across wide-ranging areas. However, extensive coverage of LEO satellites, combined with openness of channel...
详细信息
Visual grounding aims to ground an image region through natural language, which heavily relies on cross-modal alignment. Most existing methods transfer visual/linguistic knowledge separately by fully fine-tuning uni-m...
详细信息
knowledge representation learning is usually used in knowledge reasoning and other related fields. Its goal is to use low-dimensional vectors to represent the entities and relations in a knowledge graph. In the proces...
详细信息
Early classification of time series aims to accurately predict the class label of a time series as early as possible, which is significant but challenging in many time-sensitive applications. Existing early classifica...
详细信息
ISBN:
(数字)9781665488105
ISBN:
(纸本)9781665488112
Early classification of time series aims to accurately predict the class label of a time series as early as possible, which is significant but challenging in many time-sensitive applications. Existing early classification methods hold a basic closed-world assumption that the classifier must have seen the classes of test samples. However, new samples that do not belong to any trained class may appear in the real world. In this paper, we first address the early classification in an open world and design two detectors to identify which known class or unknown class a sample belongs to. Specifically, based on the observed data, an early known-class detector is designed to determine the known-class confidence and an early unknown-class detector is designed to determine the unknown-class confidence according to the Minimum Reliable Length (MRL) and the Weibull distribution of each class. Experimental results evaluated on real-world datasets demonstrate that the proposed model can identify samples of unknown and known classes accurately and early.
In practical applications, high-quality labeled data is critical for short text classification. But in many cases, it is expensive and time-consuming to obtain labeled information. Semi-supervised short text classific...
In practical applications, high-quality labeled data is critical for short text classification. But in many cases, it is expensive and time-consuming to obtain labeled information. Semi-supervised short text classification is hence attracting more attention. However, due to the sparsity of short texts, the performance of existing short text classification models always needs to be improved. Therefore, in this paper, we propose a semi-supervised Short text classification method based on Dual-Channel data Augmentation called SDCA. More specifically, in order to solve the sparsity of short texts, this model first adopts the multi-stage word-level TCN (Temporal Convolutional Network)-based attention to enhanced semantic features and an one-dimensional convolution-based attention mechanism to augment the relevance of surrounding short texts. Secondly, the unlabeled data are augmented by word embedding weighted augmentation and word replacement augmentation, so that the model can make full use of the unlabeled short texts and further enhance the network training. Finally, extensive experiments conducted on four benchmark datasets demonstrate the effectiveness of the proposed model on semi-supervised short text classification.
Investment promotion refers to the process by which the government uses disposable resources to attract investors to the region for production and business activities. The existing basic mode of attracting investment ...
详细信息
Investment promotion refers to the process by which the government uses disposable resources to attract investors to the region for production and business activities. The existing basic mode of attracting investment is to collect information about enterprises and entrepreneurs through manual methods, determine the target enterprise from the list of enterprises, and then attract investment through visits, negotiations, and other methods. As contacting and visiting companies one by one requires huge amounts of manpower and time, the choice of target companies is critical for attracting investments. However, to the best of our knowledge, no study has conducted research from the perspective of knowledge-graph-based recommendation. In this study, we define the problem of target company recommendation in the process of investment promotion, and analyze the characteristics of the problem and the challenges it faces based on the background of the actual problem. Then, a two-tier model for solving this problem is provided from the perspective of knowledge graph reasoning. Aiming at the problem that the knowledge graph will frequently change, the model is designed based on the idea of combining the advantages of global and local link prediction. The experimental results on real-world data demonstrate the effectiveness of the proposed model.
Traditional vision-based simultaneous localization and mapping (SLAM) technology cannot obtain the semantic information of the surrounding environment, which will cause the robot to fail to complete intelligent graspi...
详细信息
The remote photoplethysmography (rPPG) technique enables the estimation of vital signs such as heart rate by analyzing pulse-induced subtle skin color variation from facial videos. Robustly deriving cardiac pulse info...
详细信息
Medical image segmentation has attracted increasing attention due to its practical clinical requirements. However, the prevalence of small targets still poses great challenges for accurate segmentation. In this paper,...
详细信息
The Ambient Assisted Living (AAL) systems use sensors to detect the daily behavior of older adults and provide necessary assistance based on changes in their cognitive status and physical functions, thus enabling olde...
详细信息
暂无评论