检索结果-内蒙古大学图书馆

Proceedings of the 19th International Conference on Neural Information Processing Systems

作者： Amir Globerson Sam Roweis School of Computer Science and Engineering Interdisciplinary Center for Neural Computation The Hebrew University Jerusalem Israel Machine Learning Group Department of Computer Science University of Toronto Canada

We present an algorithm for learning a quadratic Gaussian metric (Mahalanobis distance) for use in classification tasks. Our method relies on the simple geometric intuition that a good metric is one under which points in the same class are simultaneously near each other and far from points in the other classes. We construct a convex optimization problem whose solution generates such a metric by trying to collapse all examples in the same class to a single point and push examples in other classes infinitely far away. We show that when the metric we learn is used in simple classifiers, it yields substantial improvements over standard alternatives on a variety of problems. We also discuss how the learned metric may be used to obtain a compact low dimensional feature representation of the original input space, allowing more efficient classification with very little reduction in performance.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Metamorphic electromagnetic media

Metamorphic electromagnetic media

引用

Joint 9th International Conference on Electromagnetics in Advanced Applications, ICEAA 2005 and 11th European Electromagnetic Structures Conference, EESC 2005

作者： Kyriazidou, Chryssoula A. Contopanagos, Harry F. Alexopoulos, Nicólaos G. Henry Samueli School of Engineering and Computer Science University of California Irvine Irvine CA 92697 United States Institute for Microelectronics National Center for Scientific Research Demokritos P.O. Box 60228 153 10 Aghia Paraskevi Athens Greece

ISBN: (纸本)8882020932

We analyze complex materials supplied with an integrated physical mechanism that electronically tunes their interaction with an incident electromagnetic field. We categorize the response of such scattering systems according to the values of their reflection coefficient. Characteristic ranges of these values define the metamorphic states of the system. Transitions among such states describe a fundamental change in the electromagnetic properties of the system. Under the same excitation, the metamorphic system has the built-in capability of transforming to a perfect electric conductor, a perfect absorber, a perfect electric or magnetic amplifier, a perfect magnetic conductor, etc by means of a selected switching pattern. Following the formal definition, we examine possible metamorphic properties in continuous bulk media and complex periodic metallo-dielectric structures. We find that the dispersive properties of the latter provide the necessary extra degree of freedom that allows us to obtain metamorphism. We develop specific complex media that exhibit metamorphic transitions in terms of practically realizable electronic tuning.

关键词： Magnetic amplifiers

来源：评论

学校读者我要写书评

暂无评论

Analyzing auditory neurons by learning distance functions 05

Analyzing auditory neurons by learning distance functions

引用

Proceedings of the 19th International Conference on Neural Information Processing Systems

作者： Inna Weiner Tomer Hertz Israel Nelken Daphna Weinshall School of Computer Science and Engineering The Hebrew University of Jerusalem Jerusalem Israel School of Computer Science and Engineering The Hebrew University of Jerusalem Jerusalem Israel and The Center for Neural Computation The Hebrew University of Jerusalem Jerusalem Israel The Center for Neural Computation The Hebrew University of Jerusalem Jerusalem Israel and Department of Neurobiology The Hebrew University of Jerusalem Jerusalem Israel

We present a novel approach to the characterization of complex sensory neurons. One of the main goals of characterizing sensory neurons is to characterize dimensions in stimulus space to which the neurons are highly sensitive (causing large gradients in the neural responses) or alternatively dimensions in stimulus space to which the neuronal response are invariant (defining iso-response manifolds). We formulate this problem as that of learning a geometry on stimulus space that is compatible with the neural responses: the distance between stimuli should be large when the responses they evoke are very different, and small when the responses they evoke are similar. Here we show how to successfully train such distance functions using rather limited amount of information. The data consisted of the responses of neurons in primary auditory cortex (A1) of anesthetized cats to 32 stimuli derived from natural sounds. For each neuron, a subset of all pairs of stimuli was selected such that the responses of the two stimuli in a pair were either very similar or very dissimilar. The distance function was trained to fit these constraints. The resulting distance functions generalized to predict the distances between the responses of a test stimulus and the trained stimuli.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Nearest neighbor based feature selection for regression and its application to neural activity 05

Nearest neighbor based feature selection for regression and ...

引用

Proceedings of the 19th International Conference on Neural Information Processing Systems

作者： Amir Navot Lavi Shpigelman Naftali Tishby Eilon Vaadia School of computer Science and Engineering The Hebrew University Jerusalem Israel and Interdisciplinary Center for Neural Computation The Hebrew University Jerusalem Israel Interdisciplinary Center for Neural Computation The Hebrew University Jerusalem Israel and Dept. of Physiology Hadassah Medical School The Hebrew University Jerusalem Israel

We present a non-linear, simple, yet effective, feature subset selection method for regression and use it in analyzing cortical neural activity. Our algorithm involves a feature-weighted version of the k-nearest-neighbor algorithm. It is able to capture complex dependency of the target function on its input and makes use of the leave-one-out error as a natural regularization. We explain the characteristics of our algorithm on synthetic problems and use it in the context of predicting hand velocity from spikes recorded in motor cortex of a behaving monkey. By applying feature selection we are able to improve prediction quality and suggest a novel way of exploring neural data.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Noise and the two-thirds power law 05

Noise and the two-thirds power law

引用

Proceedings of the 19th International Conference on Neural Information Processing Systems

作者： Uri Maoz Elon Portugaly Tamar Flash Yair Weiss Interdisciplinary Center for Neural Computation The Hebrew Univ. of Jerusalem Jerusalem Israel and Dept. of Computer Science and Applied Mathematics The Weizmann Inst. of Science Rehovot Israel and School of Comp. Science and Eng. The Hebrew Univ. of Jerusalem Jerusalem Israel School of Computer Science and Engineering The Hebrew University of Jerusalem Jerusalem Israel 2 Department of Computer Science and Applied Mathematics The Weizmann Institute of Science Rehovot Israel Interdisciplinary Center for Neural Computation The Hebrew University of Jerusalem Jerusalem Israel and School of Computer Science and Engineering The Hebrew University of Jerusalem Jerusalem Israel

The two-thirds power law, an empirical law stating an inverse non-linear relationship between the tangential hand speed and the curvature of its trajectory during curved motion, is widely acknowledged to be an invariant of upper-limb movement. It has also been shown to exist in eye-motion, locomotion and was even demonstrated in motion perception and prediction. This ubiquity has fostered various attempts to uncover the origins of this empirical relationship. In these it was generally attributed either to smoothness in hand- or joint-space or to the result of mechanisms that damp noise inherent in the motor system to produce the smooth trajectories evident in healthy human *** show here that white Gaussian noise also obeys this power-law. Analysis of signal and noise combinations shows that trajectories that were synthetically created not to comply with the power-law are transformed to power-law compliant ones after combination with low levels of noise. Furthermore, there exist colored noise types that drive non-power-law trajectories to power-law compliance and are not affected by smoothing. These results suggest caution when running experiments aimed at verifying the power-law or assuming its underlying existence without proper analysis of the noise. Our results could also suggest that the power-law might be derived not from smoothness or smoothness-inducing mechanisms operating on the noise inherent in our motor system but rather from the correlated noise which is inherent in this motor system.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Unified presentation contents retrieval using laser pointer information

Unified presentation contents retrieval using laser pointer ...

引用

21st International Conference on Data engineering Workshops 2005

作者： Nakano, Wataru Ochi, Yuta Kobayash, Takashi Katsuyama, Yutaka Naoi, Satoshi Yokota, Haruo Department of Computer Science Graduate School of Information Science and Engineering Tokyo Institute of Technology Ookayama 2-12-1 Meguro-ku Tokyo 152-8552 Japan Global Scientific Information and Computing Center Tokyo Institute of Technology Ookayama 2-12-1 Meguro-ku Tokyo 152-8552 Japan Fujitsu Laboratories Ltd. Kamikodanaka 4-1-1 Nakahara-ku Kanagawa 211-8588 Japan

ISBN: (纸本)0769526578

We have proposed unifying presentation contents, such as lecture video and presentation slides used in lectures, using metadata. For the unified contents, we have also proposed a search mechanism named UPRISE (Unified Presentation Slide Retrieval by Impression Search Engine). In this paper, we focus on the position, count, and duration information of a laser pointer, and propose a retrieval method using the information. In proposed method, we extract the position of laser pointer from video, and choose a sentence near the laser pointer position with a number of candidates. We evaluate our approach by applying it to actual presentation contents. © 2005 IEEE.

关键词： Information retrieval

来源：评论

学校读者我要写书评

暂无评论

Unified Presentation Contents Retrieval Using Laser Pointer Information

Unified Presentation Contents Retrieval Using Laser Pointer ...

引用

International Conference on Data engineering Workshops

作者： W. Nakano Y. Ochi T. Kobayashi Y. Katsuyama S. Naoi H. Yokota Department of Computer Science Graduate School of Information Science and Engineering Tokyo Institute of Technology Meguro Tokyo Japan Global Scientific Information and Computing Center Tokyo Institute of Technology Meguro Tokyo Japan Fujitsu Laboratories Limited Nakahara Kanagawa Japan

关键词： Information retrieval Content based retrieval Layout Laser theory Data mining Multimedia systems Feathers Electrical capacitance tomography computer science Information science

来源：评论

学校读者我要写书评

暂无评论

Direct numerical simulation of a particle-laden weak-shearing plane jet

引用

自然科学进展(英文版) 2004年第3期14卷 247-256页

作者： XIA Jun HU Guilin FAN Jianren CEN Kefa College of Computer Science and Center for Engineering and Scientific Computation Zhejiang University Hangzhou 310027 China Institute for Thermal Power Engineering College of Mechanical and Energy Engineering Zhejiang University Hangzhou 310027 China

Direct numerical simulation (DNS) has been utilized to solve numerically a two-dimensional compressible weak-shearing plane jet, in which the jet exit velocity and the co-flow velocity are in the same magnitude (2∶1 in this paper). We also use the one-way coupling method to simulate the dispersion behaviors of solid particles in various representative sizes, i.e. St=0.01, 1, 10, 100, where St is the Stokes number. We get a vorticity field with fully varicose (symmetrical) modes and the entire precise processes of the rolling-up of one spanwise vortex, the pairing of two vortexes and the mixing of three vortexes. The mean longitudinal velocity (U) profile compares well against the experimental results. The Reynolds stress profile looks special due to the symmetrical vorticity field. The particles whose St=0.01 reproduce the vortex structures in detail, and the ones of St=1 exhibit interesting self-organized behaviors and possess the most non-uniform concentration field. They disperse uniformly around the single and pairing vortex kernels, while a few of them are arranged almost in a straight line in the center region of the paring vortexes, which is caused by the contribution of the border of two vortexes in the pairing process. St=10 and 100 can be treated as large particles, on which the flow field has few impacts.

关键词： direct numerical simulation (DNS) plane turbulent jets vortex structures solid particles Stokes number

来源：评论

学校读者我要写书评

暂无评论

Passwords you'll never forget, but can't recall

Passwords you'll never forget, but can't recall

引用

Conference on Human Factors in Computing Systems, CHI EA 2004

作者： Weinshall, Daphna Kirkpatrick, Scott School of Computer Science and Engineering Center for Neural Computation Hebrew University Jerusalem Israel School of Engineering and Computer Science Hebrew University Jerusalem Israel

ISBN: (纸本)1581137036

We identify a wide range of human memory phenomena as potential certificates of identity. These "imprinting" behaviors are characterized by vast capacity for complex experiences, which can be recognized without apparent effort and yet cannot be transferred to others. They are suitable for use in near zero-knowledge protocols, which minimize the amount of secret information exposed to prying eyes while identifying an individual. We sketch several examples of such phenomena[1-3], and apply them in secure certification protocols. This provides a novel approach to human-computer interfaces, and raises new questions in several classic areas of psychology. © 2004 ACM.

关键词： Authentication

来源：评论

学校读者我要写书评

暂无评论

Learning from a small number of training examples by exploiting object categories

Learning from a small number of training examples by exploit...

引用

2004 IEEE computer Society Conference on computer Vision and Pattern Recognition Workshops, CVPRW 2004

作者： Levi, Kobi Fink, Michael Weiss, Yair School of Computer Science and Engineering Hebrew University of Jerusalem Jerusalem91904 Israel Center for Neural Computation Hebrew University of Jerusalem Jerusalem91904 Israel

In the last few years, object detection techniques have progressed immensely. Impressive detection results have been achieved for many objects such as faces [11, 14, 9] and cars [11]. The robustness of these systems emerges from a training stage utilizing thousands of positive examples. One approach to enable learning from a small set of training examples is to find an efficient set of features that accurately represent the target object. Unfortunately, automatically selecting such a feature set is a difficult task in itself. In this paper we present a novel feature selection method that is based on the notion of object categories. We assume that when learning to recognize a new object (like an apple) we also know a category it belongs to (fruit). We further assume that features that are useful for learning other objects in the same category (e.g. pear or orange) will also be useful for learning the novel object. This leads to a simple criterion for selecting features and building classifiers. We show that our method gives significant improvement in detection performance in challenging domains. © 2004 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：