458 research outputs found
Reading car license plates using deep neural networks
Abstract not availableHui Li, Peng Wang, Mingyu You, Chunhua She
Nonparametric and semiparametric regression, missing data, and related algorithms:
This dissertation consists of two chapters: Chapter 1 develops nonparametric and semiparametric regression methodologies which relate the group testing responses to the individual covariates information. In this chapter, we extend the parametric regression model of Xie (2001) for binary group testing data to the nonparametric and semiparametric models. We fit nonparametric and semiparametric models and obtain estimators of the parameters by maximizing penalized likelihood function. For implementation, we apply EM algorithm considering the individual responses as complete data and the group testing responses as observed data. Simulation studies are performed to illustrate the methodologies and to evaluate the finite sample performance of our methods. In general, group testing involves a large number of subjects, hence, the computational aspect is also discussed. The results show that our estimation methods perform well for estimating both the individual probability of positive outcome and the prevalence rate in the population.
Chapter 2 studies a partially linear regression model with missing response variable and develops semiparametric efficient inference for the parametric component of the model. The missingness considered here includes a broad range of missing patterns. For the estimation method, we use the concept of least favorable curve, least favorable direction and the generalized profile likelihood in Severini and Wong (1992). Asymptotic distributions for the estimators of the parametric components are obtained. It is shown that the estimators are asymptotically normally distributed under some conditions. Furthermore, we prove that the asymptotic covariance of the estimators achieves the semiparametric lower bound under the regularity conditions and additional conditions given in the appendix.
We also propose an algorithm which runs iteratively between fitting parametric components and fitting nonparametric components while holding the other fixed. EM algorithms are used in estimating the parametric components by a semiparametric estimating equation and in estimating the nonparametric components by smoothing methods. It is proved that the estimators from this iterative algorithm equal to the conditional expectations (conditioned on observed data) of the semiparametric efficient estimators from complete data. The methodology is illustrated and evaluated by numerical examples.Ph.D.Includes bibliographical referencesby Mingyu L
RETRACTED ARTICLE: The Notch signal mediates macrophage polarization by regulating miR-125a/miR-99b expression
We, the Editor, Publisher, and Authors of Artificial Cells, Nanomedicine and Biotechnology, have retracted the following article:Qian Li, Xia He, Qiao Yu, Yuan Wu, Mingyu Du, Jing Chen, Fanyu Peng, Wenjun Zhang, Jie Chen, Yan Wang, Hanbo Chen, Hairong Wang, Dan He & Qiang Wang. (2019). The Notch signal mediates macrophage polarization by regulating miR-125a/miR-99b expression. Artificial Cells, Nanomedicine, and Biotechnology. 47:1, 833-843, DOI: 10.1080/21691401.2019.1576711We are now cognizant that substantially similar images from this article were first published in Cancer Research. When asked, the Authors were cooperative and confirmed that the wrong images had been submitted. Further internal investigation of this article also revealed issues with the rigor of the editorial process, and therefore we have collectively agreed to retract the article.We have been informed in our decision-making by the guidance of COPE guidelines on retractionsThe retracted article will remain online to maintain the scholarly record, but it will be digitally watermarked on each page as “Retracted”
Approximation-based adaptive tracking control for MIMO nonlinear systems with input saturation
In this paper, an approximation-based adaptive tracking control approach is proposed for a class of multiinput multioutput nonlinear systems. Based on the method of neural network, a novel adaptive controller is designed via backstepping design process. Furthermore, by introducing Nussbaum function, the issue of unknown control directions is handled. In the backstepping design process, the dynamic surface control technique is employed to avoid differentiating certain nonlinear functions repeatedly. Moreover, in order to reduce the number of adaptation laws, we do not use the neural networks to directly approximate the unknown nonlinear functions but the desired control signals. Finally, we provide two examples to illustrate the effectiveness of the proposed approach.Qi Zhou, Peng Shi, Yang Tian, and Mingyu Wan
Prediction mechanisms that do not incentivize undesirable actions
A potential downside of prediction markets is that they may incentivize agents to take undesirable actions in the real world. For example, a prediction market for whether a terrorist attack will happen may incentivize terrorism, and an in-house prediction market for whether a product will be successfully released may incentivize sabotage. In this paper, we study principal-aligned prediction mechanisms–mechanisms that do not incentivize undesirable actions. We characterize all principal-aligned proper scoring rules, and we show an “overpayment” result, which roughly states that with n agents, any prediction mechanism that is principal-aligned will, in the worst case, require the principal to pay Θ(n) times as much as a mechanism that is not. We extend our model to allow uncertainties about the principal’s utility and restrictions on agents’ actions, showing a richer characterization and a similar “overpayment” result.Peng Shi, Vincent Conitzer, Mingyu Gu
Stormflow generation in a humid forest watershed controlled by antecedent wetness and rainfall amounts
Understanding the meanings and identifying the controls for the stormflow generation with complex and nonlinear behaviors is essential for the development of threshold-based hydrological theory, as well as accurate assessment and prediction for flash flood risks. However, the study of catchment emergent patterns with threelinear threshold behaviors associated with hydrological connectivity has received little attention. Therefore, utilizing soil water storage, rainfall, and streamflow data spanning 3 years in a humid forest experimental
watershed, Dujiangyan city, China, we elucidated how and where stormflow was generated with nonlinear behaviors, which were affected by antecedent wetness and rainfall amounts. Stormflow threshold behavior was taken as a function of combined gross precipitation and antecedent soil water storage, which was isolated using piecewise regression analysis with the identification of two breakpoints (i.e., generation threshold, Tg and rising threshold, Tr). It was found that the initial emergent behavior of rainfall-runoff was generally activated at the Tg, and then an abrupt shift from slow to fast flood response was possibly triggered at the Tr. These processes are important to understand the formation and development of flash floods at the watershed scale. It was noted that, above the Tr, considerably higher stormflow amounts generally occurred due to the lateral-connectivity extension of runoff contributing area from stream to neighboring hillslopes. Meanwhile, gravity-driven water movements in soil and better hydrological connectivity during the above-Tr phase readily triggered the huge flash flood disasters. The above-Tr flash floods with abrupt shifts were predominantly controlled by rainfall amounts, while initial below-Tg stormflow generation was mainly controlled by unsaturated soil water storage. More noteworthy, under heavy rainstorm conditions, the above-Tr stormflow was dominantly generated by subsurface flow, as was demonstrated at hillslope and watershed scales. These findings contribute to increasing our understanding of the controls on three-linear threshold-based hydrological behaviors, as well as of subsurface stormflow generation mechanism associated with hydrological connectivity in humid forest watersheds
[[alternative]]The Transmission of Western Learning and the Evidential Studies of Natural Knowledge in Laye Ming : The Case of Xiong Mingyu on the Generation of Hail
[[abstract]]明末耶穌會士來華傳教,引入與中國傳統不同的另類自然知識,不僅開始行程兩種文化中自然知識的競爭,也促使士人以西學為鏡來省思傳統中國的自然知識,並加以考證或考據。本文以熊明遇(1579-1649)在《則草》與《格致草》中處理冰雹生成為例,來檢視他對留存於文獻中的自然知識進行考據與反應。
筆者發現,朱熹理學在建立「格物窮理」時,將「物各有理」擴及奇異的蜥蜴生雹,肯定蜥蜴生雹有其理。此一見解從宋流傳到明末。在亞里斯多德自然則學衝擊下,熊明遇是首位關注耶穌會士對冰雹的形成與朱熹理學對蜥蜴生雹解說不一致的士人。他不但質疑儒者「雹理不明」,並將蜥蜴生雹的「奇異」貶入繆論之中,即為思索儒學(或理學)與科學進的關係提供一個異例,也為名戴自然知識的考據,提供一個堅實的佳例。熊明遇無疑是西學影響下明代自然知識考據學的先驅者。
In late Ming, while Jesuits came to propagate Christianity in China, they also transmitted alternative natural knowledge which is different from that of Chinese. This paper aims at investigating the evidential studies of traditional natural knowledge in terms of that of western learning done by Xiong Mingyu (1579-1649) in his works of Ze cao 則草 and Gezhi cao格致草.The author finds that, in Song dynasty Zhu Xi believed that hail was generated by lizards. This view became popular from Song to the end of Ming. Nevertheless, in the context of "gewu qiongli," it implies that the bizarre case of hail generated by lizards has its own reason. Under the impact of Aristotelian natural philosophy, Xiong Mingyu found that the reasons put forth to explain the generation of hail in Chinese traditions were different from that of Aristotelian philosophers. Through his evidential study on this case, he contended that the reasons offered by Confucians are not clear; especially, he considered the generation of hail by lizards to be groundless. It became a very solid case of the relation between Neo-Confucianism and science, showing that Xiong Mingyu was a pioneer of the evidential studies of natural knowledge in the encounter of western natural philosophy and Neo-Confucianism.[[fileno]]JA01_2007_p11
Reflected backward stochastic differential equations with two RCLL barriers
In this paper we consider BSDEs with Lipschitz
coefficient reflected on two discontinuous (RCLL) barriers. In this
case, we prove first the existence and uniqueness of the solution,
then we also prove the convergence of the solutions of the penalized
equations to the solution of the RBSDE. Since the method used in the
case of continuous barriers (see Cvitanic and Karatzas, Ann. Probab. 24 (1996) 2024–2056 and Lepeltier and San Martín, J. Appl. Probab. 41 (2004) 162–175) does not
work, we develop a new method, by considering the solutions of the
penalized equations as the solutions of special RBSDEs and using
some results of Peng and Xu in Annales of I.H.P. 41 (2005) 605–630
Improving Memory Access Performance of In-Memory Key-Value Store Using Data Prefetching Techniques
In-memory Key-Value stores (IMKVs) provide significantly higher performance than traditional disk-based counterparts. As memory technologies advance, IMKVs become practical for modern Big Data processing, which include financial services, e-commerce, telecommunication network, etc. Recently, various IMKVs have been proposed from both academia and industrial. In order to leverage high performance random access capability of main memory, most IMKVs employ hashing based index structures to retrieve data according to keys. Consequently, a regular memory access pattern can be observed in data retrieval from those IMKVs. Normally speaking, one access to index (hash table), which is also located in main memory, is followed by another memory access to value data. Such a regular access pattern provides a potential opportunity that data prefetching techniques can be employed to improve memory access efficiency for data retrieval in these IMKVs. Based on this observation, we explore various data prefetching techniques with proper architecture level modifications on memory controller considering trade-off between design overhead and performance. Specifically, we focus on two key design issues of prefetching techniques: (1) where to fetch data (i.e. data address)? and (2) how many data to fetch (i.e. data size)? Experimental results demonstrate that memory access performance can be substantially improved up to 35.4%. In addition, we also demonstrate the overhead of prefetching on power consumption.EICPCI-S(ISTP)[email protected]; [email protected]; [email protected]; [email protected]
Enhancing Mobility around Schiphol Airport for International Passengers
This study presents the process and results of the graduation project: Enhancing Mobility Experience around Schiphol Airport for International passengers.This project is part of this project Optimaliseren Mobiliteit rondom Schiphol (in English Optimise Mobility around Schiphol). Within its broad scope, this graduation focuses on Chinese passengers, who arrive at Schiphol Airport for the first time. It is aiming to create a care-free and seamless transport journey from/to Schiphol from their perspective. The project took place in five phases, Research&Analysis phase, Ideation phase, Conceptualisation phase, Embodiment phase and Verification phase. The Research&Analysis phase consisted of desk research, field research and analysis. The Research&Analysis phase revealed the existing ecosystem: a well-developed transport system in the Netherlands, and the potential of the Chinese outbound tourism market. This phase also showed the real experience of first-time passengers: difficulty on getting information, making a choice and feeling the sense of safety. Last but not least, it defined the design goals for the design phases: enable passengers to plan their trips based on the needs easily, and make them feel confident, supportive all the time.The Ideation phase and Conceptualisation phase developed the final concept according to the design brief established at the end of Research&Analysis phase. During this phase, several iteration cycles were conducted to develop and refine the concept. The final design is a digital platform, in the form of a WeChat Mini Program, a light application embedded in WeChat (one of the most frequently used digital product in China). It supports the users from planning their transport before leaving to conducting their transport plan after arrival by fulfilling people’s needs on choosing route, tickets, getting information and finding the right spot for boarding.In the last two phases Embodiment phase and Verification phase, a high-fidelity interactive prototype was built and tested. The user test shows the final design almost meet the design goals. Passengers feel supported during the process of choosing routes and tickets, and feel easier on completing the tasks of finding the spot and getting on the vehicle in an unfamiliar environment. However, the role of the Mini Program in a visitor's journey of visiting the Netherlands is not clearly defined. For the final design, it is recommended to accomplish the service beyond the phone screen. For Schiphol Airport and other transport operators, it is recommended to use WeChat Mini Program as the carrier of service. The final design could be a starting point of the future development of the digital platform as a Mini Program. Those conclusions from Research&Analysis stage might also be a reference for them.Design for Interactio
- …
