-
1تقرير
المؤلفون: Zhao, Zhipeng, Zhou, Kun, Wang, Xiaolei, Zhao, Wayne Xin, Pan, Fan, Cao, Zhao, Wen, Ji-Rong
مصطلحات موضوعية: Computer Science - Information Retrieval, Computer Science - Artificial Intelligence
الوصف: Conversational recommender systems (CRS) aim to provide the recommendation service via natural language conversations. To develop an effective CRS, high-quality CRS datasets are very crucial. However, existing CRS datasets suffer from the long-tail issue, \ie a large proportion of items are rarely (or even never) mentioned in the conversations, which are called long-tail items. As a result, the CRSs trained on these datasets tend to recommend frequent items, and the diversity of the recommended items would be largely reduced, making users easier to get bored. To address this issue, this paper presents \textbf{LOT-CRS}, a novel framework that focuses on simulating and utilizing a balanced CRS dataset (\ie covering all the items evenly) for improving \textbf{LO}ng-\textbf{T}ail recommendation performance of CRSs. In our approach, we design two pre-training tasks to enhance the understanding of simulated conversation for long-tail items, and adopt retrieval-augmented fine-tuning with label smoothness strategy to further improve the recommendation of long-tail items. Extensive experiments on two public CRS datasets have demonstrated the effectiveness and extensibility of our approach, especially on long-tail recommendation.
Comment: work in progressالوصول الحر: http://arxiv.org/abs/2307.11650Test
-
2تقرير
المؤلفون: Wang, Xiaolei, Zhou, Kun, Tang, Xinyu, Zhao, Wayne Xin, Pan, Fan, Cao, Zhao, Wen, Ji-Rong
مصطلحات موضوعية: Computer Science - Computation and Language, Computer Science - Information Retrieval
الوصف: Conversational recommender systems (CRSs) aim to provide recommendation services via natural language conversations. Although a number of approaches have been proposed for developing capable CRSs, they typically rely on sufficient training data for training. Since it is difficult to annotate recommendation-oriented dialogue datasets, existing CRS approaches often suffer from the issue of insufficient training due to the scarcity of training data. To address this issue, in this paper, we propose a CounterFactual data simulation approach for CRS, named CFCRS, to alleviate the issue of data scarcity in CRSs. Our approach is developed based on the framework of counterfactual data augmentation, which gradually incorporates the rewriting to the user preference from a real dialogue without interfering with the entire conversation flow. To develop our approach, we characterize user preference and organize the conversation flow by the entities involved in the dialogue, and design a multi-stage recommendation dialogue simulator based on a conversation flow language model. Under the guidance of the learned user preference and dialogue schema, the flow language model can produce reasonable, coherent conversation flows, which can be further realized into complete dialogues. Based on the simulator, we perform the intervention at the representations of the interacted entities of target users, and design an adversarial training method with a curriculum schedule that can gradually optimize the data augmentation strategy. Extensive experiments show that our approach can consistently boost the performance of several competitive CRSs, and outperform other data augmentation methods, especially when the training data is limited. Our code is publicly available at https://github.com/RUCAIBox/CFCRSTest.
Comment: Accepted by KDD 2023. Code: https://github.com/RUCAIBox/CFCRSTestالوصول الحر: http://arxiv.org/abs/2306.02842Test
-
3دورية أكاديمية
المؤلفون: Lu Mao, Kun Wang, Weiye Zhu, Zi‐cong Shen, Xian‐jun Zhang, Zhi‐yang Xie, Pan Fan, Hang Shi, Bin Zhu, Lijun Li, Guanyi Liu, Yingqing Ren, Xiao‐Tao Wu
المصدر: Orthopaedic Surgery, Vol 16, Iss 6, Pp 1336-1343 (2024)
مصطلحات موضوعية: complication, disc herniation, endoscopic spine surgery, percutaneous discectomy, Orthopedic surgery, RD701-811
الوصف: Objective The reported date in the repeat surgical intervention for adolescent lumbar disc herniation (ALDH) after percutaneous endoscopic lumbar discectomy (PELD) was quite scarce. This study aims to introduce cases of repeat surgeries after PELD for ALDH and assess the incidence, chief causes, repeat surgery methods, and surgical outcomes of repeat surgeries after PELD for ALDH. Methods A retrospective multicenter observational study was conducted on patients undergoing repeat surgeries after PELD for ALDH at four tertiary referral hospitals from January 2014 through August 2022. The incidence of repeat surgeries, chief causes, strategies for repeat surgeries, and timing of repeat surgeries were recorded and analyzed. The clinical outcomes were evaluated by the Numeric Rating Scales (NRS) scores and the modified MacNab criteria. Statistical analyses were performed with the Wilcoxon signed‐rank test. Results A total of 23 patients who underwent repeat surgeries after PELD for ALDH were included. The chief causes were re‐herniation (homo‐lateral re‐herniation at the same level, new disc herniation of adjacent level). The repeat surgery methods were revision PELD, micro‐endoscopic discectomy (MED), open discectomy and instrumented lumbar inter‐body fusion. The NRS scores decreased significantly in follow‐up evaluations and these scores demonstrated significant improvement at the last follow‐up (p
وصف الملف: electronic resource
-
4تقرير
المؤلفون: Zhang, Jinbao, Zhang, Xuan, Jiao, Lei, Granmo, Ole-Christoffer, Qian, Yongjun, Pan, Fan
مصطلحات موضوعية: Electrical Engineering and Systems Science - Signal Processing, Computer Science - Machine Learning
الوصف: Neural network-based models have found wide use in automatic long-term electrocardiogram (ECG) analysis. However, such black box models are inadequate for analysing physiological signals where credibility and interpretability are crucial. Indeed, how to make ECG analysis transparent is still an open problem. In this study, we develop a Tsetlin machine (TM) based architecture for premature ventricular contraction (PVC) identification by analysing long-term ECG signals. The architecture is transparent by describing patterns directly with logical AND rules. To validate the accuracy of our approach, we compare the TM performance with those of convolutional neural networks (CNNs). Our numerical results demonstrate that TM provides comparable performance with CNNs on the MIT-BIH database. To validate interpretability, we provide explanatory diagrams that show how TM makes the PVC identification from confirming and invalidating patterns. We argue that these are compatible with medical knowledge so that they can be readily understood and verified by a medical doctor. Accordingly, we believe this study paves the way for machine learning (ML) for ECG analysis in clinical practice.
الوصول الحر: http://arxiv.org/abs/2301.10181Test
-
5دورية أكاديمية
المؤلفون: Qing-Wei Li, Hui-Pan Fan, Li-Feng Ren, Ye-Rui Zhu, Zi-Qi Lü, Shuai-Jing Ren
المصدر: Case Studies in Thermal Engineering, Vol 55, Iss , Pp 104177- (2024)
مصطلحات موضوعية: Coal combustion, Ventilation rate, Stage characteristics, Kinetic mode, Apparent activation energy, Engineering (General). Civil engineering (General), TA1-2040
الوصف: Ventilation provides the oxygen required for coal combustion, but it also accelerates the thermal loss of reaction system. To investigate the influence of ventilation on coal combustion, the characteristic temperatures and stage variation were analyzed. Furthermore, the kinetic mode was determined and the apparent activation energy was calculated. The results indicated that the increase of ventilation rate made the boundary temperatures for different stages decreased. The contribution of thermal decomposition effect to mass loss increased, and the contribution of burning effect correspondingly decreased. The ventilation rate mainly presented an influence on the kinetic mode at the second half of coal burning. The kinetic mode transformed from three-dimensional diffusion to random nucleation and subsequent growth when the ventilation rate reached 200 ml/min. In addition, under the same conversion rate, the apparent activation energy during thermal decomposition first increased and then decreased, and reached the maximum when the ventilation rate was 100 ml/min. The apparent activation energy during coal burning first decreased and then increased, and again decreased, and the increasing trend appeared when the ventilation rate was within 100–150 ml/min. These findings will provide guidance for the control and prevention of coal combustion by adjusting air leakage.
وصف الملف: electronic resource
العلاقة: http://www.sciencedirect.com/science/article/pii/S2214157X24002089Test; https://doaj.org/toc/2214-157XTest
-
6دورية أكاديمية
المؤلفون: Chusan Zheng, Yafeng Li, Jian Li, Ning Li, Pan Fan, Jieqi Sun, Penghui Liu
المصدر: Mathematics, Vol 12, Iss 12, p 1856 (2024)
مصطلحات موضوعية: attention mechanism, convolution neural network, dynamic kernel, image classification, Mathematics, QA1-939
الوصف: Convolution is a crucial component of convolution neural networks (CNNs). However, the standard static convolution has two primary defects: data independence and the weak ability to integrate global and local features. This paper proposes a novel and efficient dynamic convolution method with global and local attention to address these issues. A building block called the Global and Local Attention Unit (GLAU) is designed, in which a weighted fusion of global channel attention kernels and local spatial attention kernels generates the proposed dynamic convolution kernels. The GLAU is data-dependent and has better adaptability and the ability to integrate global and local features into each layer. We refer to such modified CNNs with GLAUs as “GLAUNets”. Extensive evaluation experiments for image classification compared to classical CNNs and the state-of-the-art dynamic convolution neural networks were conducted on the popular benchmark datasets. In terms of classification accuracy, the number of parameters, and computational complexity, the experimental results demonstrate the outstanding performance of GLAUNets.
وصف الملف: electronic resource
-
7دورية أكاديمية
المؤلفون: Liu, Jin (ORCID
0000-0002-4242-6582 ), Burgess, Yin, DiStefano, Christine, Pan, Fan, Jiang, Ningالمصدر: Journal of Psychoeducational Assessment. Jul 2020 38(4):460-474.
تمت مراجعته من قبل الزملاء: Y
Page Count: 15
Sponsoring Agency: Institute of Education Sciences (ED)
الواصفات: Pediatrics, Symptoms (Individual Disorders), Check Lists, Preschool Children, Intervention, Psychometrics, Screening Tests, Identification, At Risk Persons, Emotional Disturbances, Behavior Disorders, Preschool Teachers, Teacher Attitudes, Guidelines, Validity, Evaluation Methods, Factor Structure, Cutting Scores, Response to Intervention, Factor Analysis, Elementary School Teachers, Goodness of Fit, Child Behavior, Rating Scales
معرفات التقييم و الدراسة: Behavior Assessment System for Children
مستخلص: In the Response to Intervention framework, a psychometrically sound screening tool is essential for identification of children with emotional and behavioral risk. The purpose of this study was to examine the validity of the Pediatric Symptom Checklist-17 (PSC-17) screener in school-based settings. Forty-four teachers rated 738 preschoolers using the PSC-17; children were later assessed using long forms of the Behavior Assessment System for Children (BASC-2) Preschool form or the Achenbach System of Empirically Based Assessment (ASEBA) Caregiver--Teacher Report Form to identify emotional and behavioral disorder. Validity evidence including examinations of a multilevel factor structure, internal consistency, and criterion-related validity supported the conclusion that the PSC-17 is a high-quality universal screening tool in school-based settings. Finally, to identify emotional and behavioral risk with young children, receiver operating characteristic curve analyses with the PSC-17 yielded a lower cutoff score (i.e., 7) than the original cutoff score (i.e., 15) based on a clinical sample.
Abstractor: As Provided
IES Funded: Yes
-
8دورية أكاديمية
المؤلفون: Su Dong, Pan Fan, Haotian Yu, Bo Jiang, Dawei Sun
المصدر: Frontiers in Endocrinology, Vol 14 (2024)
مصطلحات موضوعية: biomarker, polypoidal choroidal vasculopathy, anti-vascular endothelial growth factor, cytokine, choroidal thicknesses, pachychoroid, Diseases of the endocrine glands. Clinical endocrinology, RC648-665
الوصف: PurposePolypoidal choroidal vasculopathy (PCV) is an irreversible retinal choroidal disease. Individuals with PCV exhibit diverse baseline characteristics, including systemic characteristics, ocular traits, metabolic factor levels, and different responses to intravitreal anti-VEGF therapy. This study aims to investigate the pathogenesis of PCV by analyzing the systemic characteristics, ocular traits, and cytokine levels at baseline within a cohort of patients who exhibit different responses to anti-VEGF treatment.MethodsWe conducted a retrospective analysis involving 80 eyes diagnosed with PCV. Patients were categorized into two groups based on responses to suboptimal intravitreal ranibizumab injection therapy: those with suboptimal responses and optimal responses. Aqueous humor samples were collected from the experimental eyes, and cytokine expression levels were assessed using cytometric bead array analysis. All subjects were further stratified into two groups according to the median choroidal thickness. Subsequently, logistic regression analysis and the ROC curve were employed to examine the relationship between cytokine expression levels, choroidal thickness, and anti-VEGF response.ResultsThe results revealed that compared to the group of optimal anti-VEGF response, the choroid in the suboptimal response group exhibited a significantly greater thickness. Additionally, compared to the suboptimal anti-VEGF response group, the expression levels of VEGF and VCAM-1 were markedly lower observed in the optimal anti-VEGF response group, while TNF-α showed the opposite trend. Logistic regression analysis indicated that VEGF, VCAM-1, and TNF-α in the aqueous humor were independent risk factors for a suboptimal anti-VEGF response. After adjusting other risk factors, the risk of suboptimal anti-VEGF response decreased to 0.998-fold, 0.997-fold, and 1.294-fold. The AUC values for VEGF, VCAM-1, and TNF-α were determined to be 0.805, 0.846, and 0.897, respectively. Furthermore, the risk of VEGF, VCAM-1, and TNF-α were significantly associated with an increased risk of suboptimal anti-VEGF response after correction for risk factors in the thick choroid group.ConclusionsOur study demonstrated that PCV exhibits systemic and ocular characteristics variations based on different anti-VEGF responses. The levels of cytokines in aqueous humor were found to have a significant correlation with the anti-VEGF response in PCV. VEGF, VCAM-1, and TNF-α are potential targets for assessing treatment response in thick choroidal PCV.
وصف الملف: electronic resource
العلاقة: https://www.frontiersin.org/articles/10.3389/fendo.2023.1307337/fullTest; https://doaj.org/toc/1664-2392Test
-
9دورية أكاديمية
المؤلفون: Liu, Xinnan, Zhang, Weiqi, Han, Yichao, Cheng, Hao, Liu, Qi, Ke, Shouyu, Zhu, Fangming, Lu, Ying, Dai, Xin, Wang, Chuan, Huang, Gonghua, Su, Bing, Zou, Qiang, Li, Huabing, Zhao, Wenyi, Xiao, Lianbo, Lu, Linrong, Tong, Xuemei, Pan, Fan, Li, Hecheng, Li, Bin
المساهمون: National Natural Science Foundation of China
المصدر: Nature Communications ; volume 15, issue 1 ; ISSN 2041-1723
مصطلحات موضوعية: General Physics and Astronomy, General Biochemistry, Genetics and Molecular Biology, General Chemistry, Multidisciplinary
الوصف: Targeting tumor-infiltrating regulatory T cells (Tregs) is an efficient way to evoke an anti-tumor immune response. However, how Tregs maintain their fragility and stability remains largely unknown. IFITM3 and STAT1 are interferon-induced genes that play a positive role in the progression of tumors. Here, we showed that IFITM3-deficient Tregs blunted tumor growth by strengthening the tumor-killing response and displayed the Th1-like Treg phenotype with higher secretion of IFNγ. Mechanistically, depletion of IFITM3 enhances the translation and phosphorylation of STAT1. On the contrary, the decreased IFITM3 expression in STAT1-deficient Tregs indicates that STAT1 conversely regulates the expression of IFITM3 to form a feedback loop. Blocking the inflammatory cytokine IFNγ or directly depleting STAT1-IFITM3 axis phenocopies the restored suppressive function of tumor-infiltrating Tregs in the tumor model. Overall, our study demonstrates that the perturbation of tumor-infiltrating Tregs through the IFNγ-IFITM3-STAT1 feedback loop is essential for anti-tumor immunity and constitutes a targetable vulnerability of cancer immunotherapy.
الإتاحة: https://doi.org/10.1038/s41467-023-44391-9Test
https://www.nature.com/articles/s41467-023-44391-9.pdfTest
https://www.nature.com/articles/s41467-023-44391-9Test -
10دورية أكاديمية
المؤلفون: Pan, Fan, Liu, Qingqing
المصدر: Frontiers in Psychology ; volume 15 ; ISSN 1664-1078
الوصف: This study informed researchers about the performance of different level-specific and target-specific model fit indices in the Multilevel Latent Growth Model (MLGM) with unbalanced design. As the use of MLGMs is relatively new in applied research domain, this study helped researchers using specific model fit indices to evaluate MLGMs. Our simulation design factors included three levels of number of groups (50, 100, and 200) and three levels of unbalanced group sizes (5/15, 10/20, and 25/75), based on simulated datasets derived from a correctly specified MLGM. We evaluated the descriptive information of the model fit indices under various simulation conditions. We also conducted ANOVA to calculated the extent to which these fit indices could be influenced by different design factors. Based on the results, we made recommendations for practical and theoretical research about the fit indices. CFI- and TFI-related fit indices performed well in the MLGM and could be trustworthy to use to evaluate model fit under similar conditions found in applied settings. However, RMSEA-related fit indices, SRMR-related fit indices, and chi square-related fit indices varied by the factors included in this study and should be used with caution for evaluating model fit in the MLGM.