المؤلفون: Maaz, Muhammad, Qiao, Rui, Zhou, Yiheng, Zhang, Renxian

مصطلحات موضوعية: Computer Science - Machine Learning

الوصف: Unsupervised performance estimation, or evaluating how well models perform on unlabeled data is a difficult task. Recently, a method was proposed by Garg et al. [2022] which performs much better than previous methods. Their method relies on having a score function, satisfying certain properties, to map probability vectors outputted by the classifier to the reals, but it is an open problem which score function is best. We explore this problem by first showing that their method fundamentally relies on the ordering induced by this score function. Thus, under monotone transformations of score functions, their method yields the same estimate. Next, we show that in the binary classification setting, nearly all common score functions - the $L^\infty$ norm; the $L^2$ norm; negative entropy; and the $L^2$, $L^1$, and Jensen-Shannon distances to the uniform vector - all induce the same ordering over probability vectors. However, this does not hold for higher dimensional settings. We conduct numerous experiments on well-known NLP data sets and rigorously explore the performance of different score functions. We conclude that the $L^\infty$ norm is the most appropriate.
Comment: IJCAI 2023 Workshop on Generalizing from Limited Resources in the Open World

الوصول الحر: http://arxiv.org/abs/2306.10160Test

المؤلفون: Zhou, Yiheng¹ (AUTHOR), Tan, Ye² (AUTHOR), Zhang, Ruizhi^1,2 (AUTHOR) zhangrz1991@gmail.com, Li, Zhiguo² (AUTHOR), Chen, Han² (AUTHOR), Bai, Jingsong² (AUTHOR), Li, Lei² (AUTHOR), Shen, Qiang¹ (AUTHOR), Luo, Guoqiang^1,3 (AUTHOR) luogq@whut.edu.cn

المصدر: Journal of Applied Physics. 12/14/2023, Vol. 134 Issue 22, p1-15. 15p.

مصطلحات موضوعية: *STRESS waves, *ELASTIC wave propagation, *THEORY of wave motion, *CASCADE impactors (Meteorological instruments), *EQUATIONS of state, *LOADING & unloading

مستخلص: Quasi-isentropic loading and unloading, employing graded density impactors (GDIs) as flyers in gas gun-driven plate impact experiments, can provide novel and valuable insights into the equation of state and strength properties of the loaded material. However, the internal ballistic process may lead to spalling or debonding of the GDI due to the intricate interactions between stress waves and interfaces. In this study, the wave propagation in the GDI was analyzed using the multimaterial Lagrangian elastic-plastic model and elastic wave propagation theory. The impact of gradient direction, power-law constant p, and thickness of the first and last layers on the tensile stress was investigated. The outcomes reveal that the mechanism of generating tensile stress varies for two gradient directions. Moreover, adjusting the constant p and the layer thickness may decrease the maximum tensile stress by 74.1% (forward graded) and 95.8% (reverse graded), respectively. The outcomes of this research provide a theoretical and simulation basis for designing and fabricating GDIs to be utilized in quasi-isentropic experiments. [ABSTRACT FROM AUTHOR]

المؤلفون: Zhou, Yiheng, Tsvetkov, Yulia, Black, Alan W, Yu, Zhou

مصطلحات موضوعية: Computer Science - Computation and Language

الوصف: We study non-collaborative dialogs, where two agents have a conflict of interest but must strategically communicate to reach an agreement (e.g., negotiation). This setting poses new challenges for modeling dialog history because the dialog's outcome relies not only on the semantic intent, but also on tactics that convey the intent. We propose to model both semantic and tactic history using finite state transducers (FSTs). Unlike RNN, FSTs can explicitly represent dialog history through all the states traversed, facilitating interpretability of dialog structure. We train FSTs on a set of strategies and tactics used in negotiation dialogs. The trained FSTs show plausible tactic structure and can be generalized to other non-collaborative domains (e.g., persuasion). We evaluate the FSTs by incorporating them in an automated negotiating system that attempts to sell products and a persuasion system that persuades people to donate to a charity. Experiments show that explicitly modeling both semantic and tactic history is an effective way to improve both dialog policy planning and generation performance.
Comment: Unpublished preprint

الوصول الحر: http://arxiv.org/abs/1909.13425Test

المؤلفون: Zhou, Yiheng, He, He, Black, Alan W, Tsvetkov, Yulia

مصطلحات موضوعية: Computer Science - Computation and Language

الوصف: Negotiation is a complex activity involving strategic reasoning, persuasion, and psychology. An average person is often far from an expert in negotiation. Our goal is to assist humans to become better negotiators through a machine-in-the-loop approach that combines machine's advantage at data-driven decision-making and human's language generation ability. We consider a bargaining scenario where a seller and a buyer negotiate the price of an item for sale through a text-based dialog. Our negotiation coach monitors messages between them and recommends tactics in real time to the seller to get a better deal (e.g., "reject the proposal and propose a price", "talk about your personal experience with the product"). The best strategy and tactics largely depend on the context (e.g., the current price, the buyer's attitude). Therefore, we first identify a set of negotiation tactics, then learn to predict the best strategy and tactics in a given dialog context from a set of human-human bargaining dialogs. Evaluation on human-human dialogs shows that our coach increases the profits of the seller by almost 60%.
Comment: In Proceedings of SigDial 2019

الوصول الحر: http://arxiv.org/abs/1909.13426Test

المؤلفون: Li, Xiao, Shang, Yujing, Wei, Jiaqi, Zhou, Yiheng

المصدر: Journal of Physics: Conference Series ; volume 2634, issue 1, page 012037 ; ISSN 1742-6588 1742-6596

الوصف: Stethoscopes have an important role in non-invasive diagnosis of cardiovascular and respiratory diseases, digestive diseases, and other kinds of diseases. The emergence of high-end diagnostic devices and new diagnostic methods have caused the status of the stethoscope to decline. However, stethoscope has the advantages of simple operation, mature auscultation theory and low cost, and thus is still widely used in medical diagnosis. This paper first introduces the design and application of electronic stethoscope solutions based on contact sensors and air coupling sensors, and then introduces advanced algorithms for digital signal processing for the diagnosis and treatment of different diseases, including heart sound noise reduction algorithm, heart sound segmentation algorithm and heart sound feature extraction and recognition algorithm. Finally, this paper summarizes the application of the electronic stethoscope system in medical testing, and its future development direction. In summary, the electronic stethoscope system is a reliable medical testing tool, which can convert sound signals into digital signals through complex signal processing algorithms for more accurate detection of human physiological parameters. The research of this paper will be of great value to the research and application of electronic stethoscopes.

الإتاحة: https://doi.org/10.1088/1742-6596/2634/1/012037Test

المؤلفون: Zhou, Yiheng, Sani, Numair, Luo, Jiebo

المصدر: Special Session on Intelligent Data Mining, IEEE Big Data Conference, Washington, DC, December 2016

مصطلحات موضوعية: Computer Science - Social and Information Networks

الوصف: According to NSDUH (National Survey on Drug Use and Health), 20 million Americans consumed drugs in the past few 30 days. Combating illicit drug use is of great interest to public health and law enforcement agencies. Despite of the importance, most of the existing studies on drug uses rely on surveys. Surveys on sensitive topics such as drug use may not be answered truthfully by the people taking them. Selecting a representative sample to survey is another major challenge. In this paper, we explore the possibility of using big multimedia data, including both images and text, from social media in order to discover drug use patterns at fine granularity with respect to demographics. Instagram posts are searched and collected by drug related terms by analyzing the hashtags supplied with each post. A large and dynamic dictionary of frequent drug related slangs is used to find these posts. User demographics are extracted using robust face image analysis algorithms. These posts are then mined to find common trends with regard to the time and location they are posted, and further in terms of age and gender of the drug users. Furthermore, by studying the accounts followed by the users of drug related posts, we extract common interests shared by drug users.
Comment: IEEE Big Data Conference, Washington, DC, December 2016

الوصول الحر: http://arxiv.org/abs/1611.08351Test

المؤلفون: Zhou, Yiheng, Sani, Numair, Lee, Chia-Kuei, Luo, Jiebo

مصطلحات موضوعية: Computer Science - Social and Information Networks

الوصف: Drug use by people is on the rise and is of great interest to public health agencies and law enforcement agencies. As found by the National Survey on Drug Use and Health, 20 million Americans aged 12 years or older consumed illicit drugs in the past few 30 days. Given their ubiquity in everyday life, drug abuse related studies have received much and constant attention. However, most of the existing studies rely on surveys. Surveys present a fair number of problems because of their nature. Surveys on sensitive topics such as illicit drug use may not be answered truthfully by the people taking them. Selecting a representative sample to survey is another major challenge. In this paper, we explore the possibility of using big data from social media in order to understand illicit drug use behaviors. Instagram posts are collected using drug related terms by analyzing the hashtags supplied with each post. A large and dynamic dictionary of frequent illicit drug related slang is used to find these posts. These posts are studied to find common drug consumption behaviors with regard to time of day and week. Furthermore, by studying the accounts followed by the users of drug related posts, we hope to discover common interests shared by drug users.
Comment: 2016 International Conference on Social Computing, Behavioral-Cultural Modeling, & Prediction and Behavior Representation in Modeling and Simulation (SBP-BRiMS'16)

الوصول الحر: http://arxiv.org/abs/1604.07096Test

المؤلفون: Xiang, Dawei, Yang, Chen, Li, Hao, Zhou, Yiheng, Cao, Yueyang, Sun, Zhiwen, Xie, Qiang

المساهمون: National Natural Science Foundation of China, Natural Science Foundation of Shanghai

المصدر: IEEE Transactions on Industrial Electronics ; volume 71, issue 8, page 9800-9811 ; ISSN 0278-0046 1557-9948

مصطلحات موضوعية: Electrical and Electronic Engineering, Control and Systems Engineering

الإتاحة: https://doi.org/10.1109/tie.2023.3317850Test
http://xplorestaging.ieee.org/ielx7/41/10505837/10275779.pdf?arnumber=10275779Test

المؤلفون: ZHOU Yiheng, YAN Jiaming, WU Xinzhong, REN Zihui

المصدر: Gong-kuang zidonghua, Vol 46, Iss 11, Pp 34-40 (2020)

مصطلحات موضوعية: mine power transformer, ordinary power transformers, aging of insulation paper, microstructure, electrical property, Mining engineering. Metallurgy, TN1-997

الوصف: Collect screen paperboard of mine power transformer and ordinary power transformer in practical lifetime cycle to make insulation test samples, and according to the standard of power, use test means such as viscosity tester, scanning electron microscopy to compare and analyze microstructures of insulation paper such as polymerization degree, surface morphology of mine power transformer and ordinary power transformer, as well as electrical properties such as partial discharge and electrical strength; at the same time, effect of microstructure changes on electrical performances during aging was discussed. The analysis results show that the insulation defects gradually form and expand in the insulation paper during the aging process; compared with the insulation paper of ordinary power transformers, the surface morphology changes of the insulation paper of mine power transformers are more obvious, and the formed defects are more significant.In the aging process, the polymerization degree of insulation paper of ordinary power transformer and mine power transformer presents a gradual decline trend, but the decline speed of polymerization degree of mine power transformer insulation paper is faster.The operating environment of mining area is the main reason that the aging of insulation paper of mine power transformer is more serious than that of ordinary power transformer.In the process of operation, the insulation paper of mine power transformer will produce more discharge quantity, more discharge times and wider discharge phase than ordinary power transformer.In the aging process, the electrical strength of insulation paper of both ordinary power transformer and mine power transformer decreases gradually, but the electrical strength of insulation paper of mine power transformer is lower than that of ordinary power transformer.The microstructural defects of insulation paper of mine power transformers are the direct causes of the more significant changes in their electrical properties.In the process of state monitoring and maintenance of power equipment, it is suggested that the insulation state monitoring of mine power transformers should be carried out strictly according to the power standards,electrical property parameters are more suitable for insulation status monitoring and fault diagnosis of mine power transformer.

وصف الملف: electronic resource

العلاقة: https://doaj.org/toc/1671-251XTest

الوصول الحر: https://doaj.org/article/95911a46ad76472aaae1584876d60954Test

المؤلفون: Zhou, Yiheng^1,2 (AUTHOR) zhouyiheng@semi.ac.cn, Ma, Kainan¹ (AUTHOR) makainan@semi.ac.cn, Sun, Qian¹ (AUTHOR) sunqian@semi.ac.cn, Wang, Zhaoyuxuan¹ (AUTHOR) wangzhaoyuxuan@semi.ac.cn, Liu, Ming^1,2 (AUTHOR) liuming@semi.ac.cn

المصدر: Information (2078-2489). Apr2024, Vol. 15 Issue 4, p198. 13p.

مصطلحات موضوعية: *ARTIFICIAL neural networks, *CONVOLUTIONAL neural networks, *TRANSFORMER models

مستخلص: Over the past several decades, deep neural networks have been extensively applied to medical image segmentation tasks, achieving significant success. However, the effectiveness of traditional deep segmentation networks is substantially limited by the small scale of medical datasets, a limitation directly stemming from current medical data acquisition capabilities. To this end, we introduce AttEUnet, a medical cell segmentation network enhanced by edge attention, based on the Attention U-Net architecture. It incorporates a detection branch enhanced with edge attention and a learnable fusion gate unit to improve segmentation accuracy and convergence speed on small medical datasets. The AttEUnet allows for the integration of various types of prior information into the backbone network according to different tasks, offering notable flexibility and generalization ability. This method was trained and validated on two public datasets, MoNuSeg and PanNuke. The results show that AttEUnet significantly improves segmentation performance on small medical datasets, especially in capturing edge details, with F1 scores of 0.859 and 0.888 and Intersection over Union (IoU) scores of 0.758 and 0.794 on the respective datasets, outperforming both convolutional neural networks (CNNs) and transformer-based baseline networks. Furthermore, the proposed method demonstrated a convergence speed over 10.6 times faster than that of the baseline networks. The edge attention branch proposed in this study can also be added as an independent module to other classic network structures and can integrate more attention priors based on the task at hand, offering considerable scalability. [ABSTRACT FROM AUTHOR]