Search

Your search keyword '"Shen, Xiaowei"' showing total 535 results
Start Over Author "Shen, Xiaowei"
535 results on '"Shen, Xiaowei"'

18. CAN: Constrained Attention Networks for Multi-Aspect Sentiment Analysis

Author

Hu, Mengting, Zhao, Shiwan, Zhang, Li, Cai, Keke, Su, Zhong, Cheng, Renhong, and Shen, Xiaowei

Subjects
Computer Science - Computation and Language
Abstract

Aspect level sentiment classification is a fine-grained sentiment analysis task. To detect the sentiment towards a particular aspect in a sentence, previous studies have developed various attention-based methods for generating aspect-specific sentence representations. However, the attention may inherently introduce noise and downgrade the performance. In this paper, we propose constrained attention networks (CAN), a simple yet effective solution, to regularize the attention for multi-aspect sentiment analysis, which alleviates the drawback of the attention mechanism. Specifically, we introduce orthogonal regularization on multiple aspects and sparse regularization on each single aspect. Experimental results on two public datasets demonstrate the effectiveness of our approach. We further extend our approach to multi-task settings and outperform the state-of-the-art methods., Comment: Accepted by EMNLP 2019

Published
2018

25. Hierarchical Prototype-Aligned Graph Neural Network for Cross-Scene Hyperspectral Image Classification.

Author

Shen, Danyao, Hu, Haojie, He, Fang, Zhang, Fenggan, Zhao, Jianwei, and Shen, Xiaowei

Subjects
GRAPH neural networks, IMAGE recognition (Computer vision), DATA structures, DISTRIBUTION (Probability theory), LANDSAT satellites
Abstract

The objective of cross-scene hyperspectral image (HSI) classification is to develop models capable of adapting to the "domain gap" that exists between different scenes, enabling accurate object classification in previously unseen scenes. Many researchers have devised various domain adaptation techniques aimed at aligning the statistical or spectral distributions of data from diverse scenes. However, many previous studies have overlooked the potential benefits of incorporating spatial topological information from hyperspectral imagery, which could provide a more accurate representation of the inherent data structure in HSIs. To overcome this issue, we introduce an innovative approach for cross-scene HSI classification, founded on hierarchical prototype graph alignment. Specifically, this method leverages prototypes as representative embedded representations of all samples within the same class. By employing multiple graph convolution and pooling operations, multi-scale domain alignment is attained. Beyond statistical distribution alignment, we integrate graph matching to effectively reconcile semantic and topological information. Experimental results on several datasets achieve significantly improved accuracy and generalization capabilities for cross-scene HSI classification tasks. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

39. Appropriateness of using patient-derived xenograft models for pharmacologic evaluation of novel therapies for esophageal/gastro-esophageal junction cancers.

Author

Dodbiba, Lorin, Teichman, Jennifer, Fleet, Andrew, Thai, Henry, Starmans, Maud HW, Navab, Roya, Chen, Zhuo, Girgis, Hala, Eng, Lawson, Espin-Garcia, Osvaldo, Shen, Xiaowei, Bandarchi, Bizhan, Schwock, Joerg, Tsao, Ming-Sound, El-Zimaity, Hala, Der, Sandy D, Xu, Wei, Bristow, Robert G, Darling, Gail E, Boutros, Paul C, Ailles, Laurie E, and Liu, Geoffrey

Subjects
Esophagogastric Junction, Animals, Mice, Inbred NOD, Humans, Mice, Mice, SCID, Esophageal Neoplasms, Stomach Neoplasms, Xenograft Model Antitumor Assays, Inbred NOD, SCID, General Science & Technology
Abstract

The high morbidity and mortality of patients with esophageal (E) and gastro-esophageal junction (GEJ) cancers, warrants new pre-clinical models for drug testing. The utility of primary tumor xenografts (PTXGs) as pre-clinical models was assessed. Clinicopathological, immunohistochemical markers (p53, p16, Ki-67, Her-2/neu and EGFR), and global mRNA abundance profiles were evaluated to determine selection biases of samples implanted or engrafted, compared with the underlying population. Nine primary E/GEJ adenocarcinoma xenograft lines were further characterized for the spectrum and stability of gene/protein expression over passages. Seven primary esophageal adenocarcinoma xenograft lines were treated with individual or combination chemotherapy. Tumors that were implanted (n=55) in NOD/SCID mice had features suggestive of more aggressive biology than tumors that were never implanted (n=32). Of those implanted, 21/55 engrafted; engraftment was associated with poorly differentiated tumors (p=0.04) and older patients (p=0.01). Expression of immunohistochemical markers were similar between patient sample and corresponding xenograft. mRNA differences observed between patient tumors and first passage xenografts were largely due to loss of human stroma in xenografts. mRNA patterns of early vs late passage xenografts and of small vs large tumors of the same passage were similar. Complete resistance was present in 2/7 xenografts while the remaining tumors showed varying degrees of sensitivity, that remained constant across passages. Because of their ability to recapitulate primary tumor characteristics during engraftment and across serial passaging, PTXGs can be useful clinical systems for assessment of drug sensitivity of human E/GEJ cancers.

Published
2015

45. Patterns, perceptions, and perceived barriers to physical activity in adult cancer survivors

Author

Eng, Lawson, Pringle, Dan, Su, Jie, Shen, XiaoWei, Mahler, Mary, Niu, Chongya, Charow, Rebecca, Tiessen, Kyoko, Lam, Christine, Halytskyy, Oleksandr, Naik, Hiten, Hon, Henrique, Irwin, Margaret, Pat, Vivien, Gonos, Christina, Chan, Catherine, Villeneuve, Jodie, Harland, Luke, Shani, Ravi M., Brown, M. Catherine, Selby, Peter, Howell, Doris, Xu, Wei, Liu, Geoffrey, Alibhai, Shabbir M. H., and Jones, Jennifer M.

Published
2018
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results