检索结果-内蒙古大学图书馆

Lecture Notes in computer science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 2014年 8881卷 283-298页

作者： Kanj, Iyad Lin, Guohui Liu, Tian Tong, Weitian Xia, G. Xu, Jinhui Yang, Boting Zhang, Fenghui Zhang, Peng Zhu, Binhai School of Computing DePaul University 243 S. Wabash Avenue ChicagoIL60604 United States Department of Computing Science University of Alberta EdmontonABT6G 2E8 Canada School of Electronic Engineering and Computer Science Peking University Beijing100871 China Department of Computing Science Lafayette college EastonPA18042 United States 338 Davis Hall BuffaloNY14260 United States Department of Computer Science University of Regina ReginaSKS4S 0A2 Canada Google Kirkland 747 6th Street South KirklandWA98033 United States School of Computer Science and Technology Shandong University Jinan250101 China Department of Computer Science Montana State University BozemanMT59717 United States

We study the multicut on trees and the generalized multiway cut on trees problems. For the multicut on trees problem, we present a parameterized algorithm that runs in time O∗ (ρk), where ρ = _ √ 2 + 1 ≈ 1.555 is the positive root of the polynomial x4 - 2x2 - 1. This improves the current-best algorithm of Chen et al. that runs in time O∗ (1.619k). By reducing generalized multiway cut on trees to multicut on trees, our results give a parameterized algorithm that solves the generalized multiway cut on trees problem in time O∗ (ρk). We also show that the generalized multiway cut on trees problem is solvable in polynomial time if the number of terminal sets is a fixed constant. Springer International Publishing Switzerland 2014.

关键词： Polynomial approximation

来源：评论

学校读者我要写书评

暂无评论

Distributed hypothesis testing with social learning and symmetric fusion

Distributed hypothesis testing with social learning and symm...

引用

作者： Rhim, Joong Bum Goyal, Vivek K Department of Electrical Engineering and Computer Science Research Laboratory of Electronics Massachusetts Institute of Technology CambridgeMA02139 United States GroupM Data and Analytics New YorkNY10001 United States Boston University Massachusetts Institute of Technology United States Google United States Department of Electrical and Computer Engineering Boston University BostonMA02215 United States

We study the utility of social learning in a distributed detection model with agents sharing the same goal: a collective decision that optimizes an agreed upon criterion. We show that social learning is helpful in some cases but is provably futile (and thus essentially a distraction) in other cases. Specifically, we consider Bayesian binary hypothesis testing performed by a distributed detection and fusion system, where all decision-making agents have binary votes that carry equal weight. Decision-making agents in the team sequentially make local decisions based on their own private signals and all precedent local decisions. It is shown that the optimal decision rule is not affected by precedent local decisions when all agents observe conditionally independent and identically distributed private signals. Perfect Bayesian reasoning will cancel out all effects of social learning. When the agents observe private signals with different signal-to-noise ratios, social learning is again futile if the team decision is only approved by unanimity. Otherwise, social learning can strictly improve the team performance. Furthermore, the order in which agents make their decisions affects the team decision. © 1991-2012 IEEE.

关键词： Signal to noise ratio

来源：评论

学校读者我要写书评

暂无评论

Exploiting group recommendation functions for flexible preferences

Exploiting group recommendation functions for flexible prefe...

引用

International Conference on Data Engineering

作者： Senjuti Basu Roy Saravanan Thirumuruganathan Sihem Amer-Yahia Gautam Das Cong Yu Institute of Technology University of Washington Tacoma Computer Science Department University of Texas Arlington CNRS-LIG Google Research

We examine the problem of enabling the flexibility of updating one's preferences in group recommendation. In our setting, any group member can provide a vector of preferences that, in addition to past preferences and other group members' preferences, will be accounted for in computing group recommendation. This functionality is essential in many group recommendation applications, such as travel planning, online games, book clubs, or strategic voting, as it has been previously shown that user preferences may vary depending on mood, context, and company (i.e., other people in the group). Preferences are enforced in an feedback box that replaces preferences provided by the users by a potentially different feedback vector that is better suited for maximizing the individual satisfaction when computing the group recommendation. The feedback box interacts with a traditional recommendation box that implements a group consensus semantics in the form of Aggregated Voting or Least Misery, two popular aggregation functions for group recommendation. We develop efficient algorithms to compute robust group recommendations that are appropriate in situations where users have changing preferences. Our extensive empirical study on real world data-sets validates our findings.

关键词： Vectors Hamming distance Robustness Poles and towers Complexity theory Approximation algorithms Semantics

来源：评论

学校读者我要写书评

暂无评论

Input output kernel regression: supervised and semi-supervised structured output prediction with operator-valued kernels

The Journal of Machine Learning Research

引用

The Journal of Machine Learning Research 2016年第1期17卷

作者： Kevin Murphy Bernhard Schölkopf Céline Brouard Marie Szafranski Florence D'Alché-Buc Google MPI for Intelligent Systems Helsinki Institute for Information Technology Department of Computer Science Aalto University Espoo Finland and IBISC Université d'Évry Val d'Essonne Évry Cedex France ENSIIE & LaMME Université d'Évry Val d'Essonne CNRS INRA Évry Cedex France and IBISC Université d'Évry Val d'Essonne Évry Cedex France LTCI CNRS Télécom ParisTech Université Paris-Saclay Paris France and IBISC Université d'Évry Val d'Essonne Évry Cedex France

In this paper, we introduce a novel approach, called Input Output Kernel Regression (IOKR), for learning mappings between structured inputs and structured outputs. The approach belongs to the family of Output Kernel Regression methods devoted to regression in feature space endowed with some output kernel. In order to take into account structure in input data and benefit from kernels in the input space as well, we use the Reproducing Kernel Hilbert Space theory for vector-valued functions. We first recall the ridge solution for supervised learning and then study the regularized hinge loss-based solution used in Maximum Margin Regression. Both models are also developed in the context of semi-supervised setting. In addition we derive an extension of Generalized Cross Validation for model selection in the case of the least-square model. Finally we show the versatility of the IOKR framework on two different problems: link prediction seen as a structured output problem and multi-task regression seen as a multiple and interdependent output problem. Eventually, we present a set of detailed numerical results that shows the relevance of the method on these two tasks.

关键词： operator-valued kernel output kernel regression semi-supervised learning structured output prediction vector-valued RKHS

来源：评论

学校读者我要写书评

暂无评论

Programmable in-loop deblock filter processor for video decoders

Programmable in-loop deblock filter processor for video deco...

引用

IEEE Workshop on Signal Processing Systems (SIPS)

作者： Janne Janhunen Pekka Jääskeläinen Jari Hannuksela Tero Rintaluoma Aki Kuusela Centre for Wireless Communications University of Oulu Department of Pervasive Computing Tampere University of Technology Computer Science and Engineering department University of Oulu Google Finland

The short time to market cycle and the target to reduce design and verification costs are driving forces to design programmable implementations of the video processing algorithms. We present two processor architectures the first one representing an application-specific instruction set processor (ASIP) design, whereas the second architecture represents a domain-specific instruction-set processor (DSIP) architecture with more general purpose instruction-set. In this work, we present results for H264 and VP8 in-loop deblocking algorithms. The processors are based on the transport triggered architecture which provides scalable instruction-level parallelism and, thanks to its simple structure, lend itself to cost effective designs. Both of the designs are programmed with C language with a minimal additional parallelism markup. The designs fulfill realtime requirements for filtering macroblocks in high-definition video. The first architecture, based on special function units, filters a high-definition stream (1920 × 1080) at 75 fps, whereas the second architecture, which provides a better programmability, filters the stream at 53 fps. The processors run on 200 MHz clock frequency and the areas vary from 146k to 373k gate equivalents depending on the processor architecture.

关键词： computer architecture Clocks Parallel processing Hardware Algorithm design and analysis Complexity theory Vectors

来源：评论

学校读者我要写书评

暂无评论

Endurance-aware cache line management for non-volatile caches

Endurance-aware cache line management for non-volatile cache...

引用

作者： Wang, Jue Dong, Xiangyu Xie, Yuan Jouppi, Norman P. Computer Science and Engineering Department Pennsylvania State University 111N 1st Building University Park PA 16802 United States Qualcomm Technology 5775 Morehouse Dr San Diego CA 92121 United States Google Inc. 1600 Amphitheatre Parkway Mountain View CA 94043 United States

Nonvolatile memories (NVMs) have the potential to replace low-level SRAM or eDRAM on-chip caches because NVMs save standby power and provide large cache capacity. However, limited write endurance is a common problem for NVM technologies, and today's cache management might result in unbalanced cache write traffic, causing heavily written cache blocks to fail much earlier than others. Although wear-leveling techniques for NVM-based main memories exist, we cannot simply apply them to NVM-based caches. This is because cache writes have intraset variations as well as interset variations, while writes to main memories only have interset variations. To solve this problem, we propose i 2WAP, a new cache management policy that can reduce both inter-and intraset write variations. i2WAP has two features: Swap-Shift, an enhancement based on existing main memory wear leveling to reduce cache interset write variations, and Probabilistic Set Line Flush, a novel technique to reduce cache intraset write variations. Implementing i2WAP only needs two global counters and two global registers. In one of our studies, i 2WAP can improve the NVM cache lifetime by 75% on average and up to 224%. We also validate that i2WAP is effective in systems with different cache configurations and workloads. © 2014 ACM.

关键词： Memory architecture

来源：评论

学校读者我要写书评

暂无评论

Immutability

Lecture Notes in Computer Science (including subseries Lectu...

引用

Lecture Notes in computer science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 2013年 7850卷 233-269页

作者： Potanin, Alex Östlund, Johan Zibin, Yoav Ernst, Michael D. School of Engineering and Computer Science VUW Wellington New Zealand Department of Information Technology Uppsala University Uppsala Sweden Google New York NY United States Computer Science and Engineering University of Washington WA United States

ISBN: (纸本)9783642369452

One of the main reasons aliasing has to be controlled, as highlighted in another chapter [1] of this book [2], is the possibility that a variable can unexpectedly change its value without the referrer's knowledge. This book will not be complete without a discussion of the impact of immutability on reference-abundant imperative object-oriented languages. In this chapter we briefly survey possible definitions of immutability and present recent work by the authors on adding immutability to object-oriented languages and how it impacts aliasing. © Springer-Verlag Berlin Heidelberg 2013.

关键词： Object oriented programming

来源：评论

学校读者我要写书评

暂无评论

Cover tree Bayesian reinforcement learning

The Journal of Machine Learning Research

引用

The Journal of Machine Learning Research 2014年第1期15卷

作者： Kevin Murphy Bernhard Schölkopf Nikolaos Tziortziotis Christos Dimitrakakis Konstantinos Blekas Google Department of Computer Science and Engineering University of Ioannina Greece Department of Computer Science and Engineering Chalmers University of Technology Sweden

This paper proposes an online tree-based Bayesian approach for reinforcement learning. For inference, we employ a generalised context tree model. This defines a distribution on multivariate Gaussian piecewise-linear models, which can be updated in closed form. The tree structure itself is constructed using the cover tree method, which remains efficient in high dimensional spaces. We combine the model with Thompson sampling and approximate dynamic programming to obtain effective exploration policies in unknown environments. The flexibility and computational simplicity of the model render it suitable for many reinforcement learning problems in continuous state spaces. We demonstrate this in an experimental comparison with a Gaussian process model, a linear model and simple least squares policy iteration.

关键词： Bayesian inference non-parametric statistics reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Deceptive answer prediction with user preference graph

Deceptive answer prediction with user preference graph

引用

51st Annual Meeting of the Association for Computational Linguistics, ACL 2013

作者： Li, Fangtao Gao, Yang Zhou, Shuchang Si, Xiance Dai, Decheng Google Research Mountain View Institute of Computing Technology CAS China State Key Laboratory of Computer Architecture Institute of Computing Technology CAS China Department of Computer Science and Technology Tsinghua University China

ISBN: (纸本)9781937284503

In Community question answering (QA) sites, malicious users may provide deceptive answers to promote their products or services. It is important to identify and filter out these deceptive answers. In this paper, we first solve this problem with the traditional supervised learning methods. Two kinds of features, including textual and contextual features, are investigated for this task. We further propose to exploit the user relationships to identify the deceptive answers, based on the hypothesis that similar users will have similar behaviors to post deceptive or authentic answers. To measure the user similarity, we propose a new user preference graph based on the answer preference expressed by users, such as "helpful" voting and "best answer" selection. The user preference graph is incorporated into traditional supervised learning framework with the graph regularization technique. The experiment results demonstrate that the user preference graph can indeed help improve the performance of deceptive answer prediction. © 2013 Association for Computational Linguistics.

关键词： Supervised learning

来源：评论

学校读者我要写书评

暂无评论

Gibbs max-margin topic models with data augmentation

The Journal of Machine Learning Research

引用

The Journal of Machine Learning Research 2014年第1期15卷

作者： Kevin Murphy Bernhard Schölkopf Jun Zhu Ning Chen Hugh Perkins Bo Zhang Google State Key Lab of Intelligent Technology and Systems Tsinghua National Lab for Information Science and Technology Department of Computer Science and Technology Tsinghua University Beijing China

Max-margin learning is a powerful approach to building classifiers and structured output predictors. Recent work on max-margin supervised topic models has successfully integrated it with Bayesian topic models to discover discriminative latent semantic structures and make accurate predictions for unseen testing data. However, the resulting learning problems are usually hard to solve because of the non-smoothness of the margin loss. Existing approaches to building max-margin supervised topic models rely on an iterative procedure to solve multiple latent SVM subproblems with additional mean-field assumptions on the desired posterior distributions. This paper presents an alternative approach by defining a new max-margin loss. Namely, we present Gibbs max-margin supervised topic models, a latent variable Gibbs classifier to discover hidden topic representations for various tasks, including classification, regression and multi-task learning. Gibbs max-margin supervised topic models minimize an expected margin loss, which is an upper bound of the existing margin loss derived from an expected prediction rule. By introducing augmented variables and integrating out the Dirichlet variables analytically by conjugacy, we develop simple Gibbs sampling algorithms with no restrictive assumptions and no need to solve SVM subproblems. Furthermore, each step of the "augment-and-collapse" Gibbs sampling algorithms has an analytical conditional distribution, from which samples can be easily drawn. Experimental results on several medium-sized and large-scale data sets demonstrate significant improvements on time effciency. The classification performance is also improved over competitors on binary, multi-class and multi-label classification tasks.

关键词： Gibbs classifiers max-margin learning regularized Bayesian inference supervised topic models support vector machines

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：