检索结果-内蒙古大学图书馆

2006 IEEE International Conference on Image Processing, ICIP 2006

作者： Sahba, Farhang Tizhoosh, Hamid R. Salama, Magdy M. A. Systems Design Engineering Pattern Analysis and Machine Intelligence Laboratory University of Waterloo Waterloo ON Canada Electrical and Computer Engineering Pattern Analysis and Machine Intelligence Laboratory University of Waterloo 200 University Ave W Waterloo ON Canada

ISBN: (纸本)1424404819

In this paper a new approach to object extraction and recognition based on reinforcement learning is presented. We use this novel idea as a method to optimally segment the image and increase the recognition rate. The success rate is compared with a classical approach. Preliminary results demonstrate increase in recognition rate. ©2006 IEEE.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Improving the convergence of backpropagation by opposite transfer functions

Improving the convergence of backpropagation by opposite tra...

引用

IEEE International Joint Conference on Neural Network

作者： Ventresca, Mario Tizhoosh, Hamid R. Pattern Analysis and Machine Intelligence (PAMI) Laboratory Systems Design Engineering Department University of Waterloo Waterloo ON N2L 3G1 Canada

ISBN: (纸本)9780780394902

The backpropagation algorithm is a very, popular approach to learning in feed-forward multi-layer perceptron networks. However, in many scenarios the time required to adequately learn the task is considerable. Many existing approaches have improved the convergence rate by altering the learning algorithm. We present a simple alternative approach inspired by opposition-based learning that simultaneously considers each network transfer function and its opposite. The effect is an improvement in convergence rate and over traditional backpropagation learning with momentum. We use four common benchmark problems to illustrate the improvement in convergence time.

关键词： Multilayer neural networks

来源：评论

学校读者我要写书评

暂无评论

Opposition-based Q(λ) algorithm

Opposition-based Q(λ) algorithm

引用

International Joint Conference on Neural Networks 2006, IJCNN '06

作者： Shokri, Maryam Tizhoosh, Hamid R. Kamel, Mohamed Pattern Analysis and Machine Intelligence Laboratory Department of Systems Design Engineering University of Waterloo 200 University Avenue West ON N2L 3G1 Canada Pattern Analysis and Machine Intelligence Laboratory Department of Electrical and Computer Engineering University of Waterloo 200 University Avenue West ON N2L 3G1 Canada

ISBN: (纸本)0780394909

The problem of delayed reward in reinforcement learning is usually tackled by implementing the mechanism of eligibility traces. In this paper we introduce an extension of eligibility traces to solve one of the challenging problems in reinforcement learning. The concept of opposition traces is proposed in this work to deal with large state space problems in reinforcement learning applications. We combine the idea of opposition and eligibility traces to construct the oppositionbased Q(λ). The results are compared with the conventional Watkins' Q(λ) and reflect a remarkable performance increase. © 2006 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Modular-Based Classifier for Phoneme Recognition

Modular-Based Classifier for Phoneme Recognition

引用

IEEE International Symposium on Signal Processing and Information Technology (ISSPIT)

作者： Abbas Ahmadi Fakhri Karray Mohamed Kamel Systems Design Engineering Department Pattern Analysis and Machine Intelligence Laboratory Waterloo ONT Canada Electrical and Computer Engineering Department Pattern Analysis and Machine Intelligence Laboratory Waterloo ONT Canada

This paper proposes a modular-based classifier for the problem of phoneme recognition. This is carried out by the use of a two-level classification approach including, high and low levels. We propose a new concept called phoneme family. To obtain phoneme families, we employ k-mean clustering method. A given unknown phoneme is first classified into a phoneme family at high level classification. Then, the exact label of the phoneme is determined at low level classification. We have used a combined framework of statistical and neural network based classifiers. Encouraging results are obtained by applying the proposed method on TIMIT database and its performance is compared against other methods

关键词： Neural networks Speech recognition Recurrent neural networks pattern recognition design engineering Databases Network topology Signal processing Information technology pattern analysis

来源：评论

学校读者我要写书评

暂无评论

Opposition-Based Q(λ) Algorithm

Opposition-Based Q(λ) Algorithm

引用

International Joint Conference on Neural Networks (IJCNN)

作者： M. Shokri H.R. Tizhoosh M. Kamel Pattern Analysis and Machine Intelligence Laboratory Department of Systems Design Engineering University of Waterloo ONT Canada Pattern Analysis and Machine Intelligence Laboratory Department of Electrical and Computer Engineering University of Waterloo ONT Canada

The problem of delayed reward in reinforcement learning is usually tackled by implementing the mechanism of eligibility traces. In this paper we introduce an extension of eligibility traces to solve one of the challenging problems in reinforcement learning. The concept of opposition traces is proposed in this work to deal with large state space problems in reinforcement learning applications. We combine the idea of opposition and eligibility traces to construct the opposition-based Q(lambda). The results are compared with the conventional Watkins' Q(lambda) and reflect a remarkable performance increase.

关键词： Learning Delay Acceleration Neural networks pattern analysis machine intelligence Laboratories System analysis and design design engineering systems engineering and theory

来源：评论

学校读者我要写书评

暂无评论

Improving the Convergence of Backpropagation by Opposite Transfer Functions

Improving the Convergence of Backpropagation by Opposite Tra...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： M. Ventresca H.R. Tizhoosh Pattern Analysis and Machine Intelligence (PAMI) laboratory in the Systems Design Engineering Department University of Waterloo Waterloo ONT Canada

The backpropagation algorithm is a very popular approach to learning in feed-forward multi-layer perceptron networks. However, in many scenarios the time required to adequately learn the task is considerable. Many existing approaches have improved the convergence rate by altering the learning algorithm. We present a simple alternative approach inspired by opposition-based learning that simultaneously considers each network transfer function and its opposite. The effect is an improvement in convergence rate and over traditional backpropagation learning with momentum. We use four common benchmark problems to illustrate the improvement in convergence time.

关键词： Convergence Transfer functions Neural networks Backpropagation algorithms Genetic algorithms pattern analysis machine intelligence Laboratories System analysis and design design engineering

来源：评论

学校读者我要写书评

暂无评论

Increasing Object Recognition Rate using Reinforced Segmentation

Increasing Object Recognition Rate using Reinforced Segmenta...

引用

IEEE International Conference on Image Processing

作者： Farhang Sahba Hamid R. Tizhoosh Magdy M.A. Salama Systems Design Engineering University of Waterloo Waterloo ONT Canada Electrical and Computer Engineering Pattern Analysis and Machine Intelligence Laboratory Waterloo ONT Canada

关键词： Object recognition Learning Image segmentation Data mining design engineering pattern analysis machine intelligence Laboratories Image recognition Object detection

来源：评论

学校读者我要写书评

暂无评论

Search Space analysis of Recurrent Spiking and Continuous-time Neural Networks

Search Space Analysis of Recurrent Spiking and Continuous-ti...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： M. Ventresca B. Ombuki Systems Design Engineering and member of the Pattern Analysis and Machine Intelligence Laboratory University of Waterloo Waterloo ONT Canada Department of Computing and Information Science University of Guelph Guelph ONT Canada Computer Science Department Brock University Saint Catharines ONT Canada

The problem of designing recurrent continuous-time and spiking neural networks is NP-Hard. A common practice is to utilize stochastic searches, such as evolutionary algorithms, to automatically construct acceptable networks. The outcome of the stochastic search is related to its ability to navigate the search space of neural networks and discover those of high quality. In this paper we investigate the search space associated with designing the above recurrent neural networks in order to differentiate which network should be easier to automatically design via a stochastic search. Our investigation utilizes two popular dynamic systems problems; (1) the Henon map and (2) the inverted pendulum as a benchmark.

关键词： Neural networks Recurrent neural networks Stochastic processes Chaotic communication Neurons Biological system modeling Evolutionary computation Navigation Helium Visualization

来源：评论

学校读者我要写书评

暂无评论

Opposition-Based Learning: A New Scheme for machine intelligence

Opposition-Based Learning: A New Scheme for Machine Intellig...

引用

International Conference on Computational intelligence for Modelling, Control and Automation, and International Conference on Intelligent Agents, Web Technologies and Internet Commerce

作者： H.R. Tizhoosh Pattern Analysis and Machine Intelligence Lab Systems Design Engineering University of Waterloo Waterloo ONT Canada

Opposition-based learning as a new scheme for machine intelligence is introduced. Estimates and counter-estimates, weights and opposite weights, and actions versus counter-actions are the foundation of this new approach. Examples are provided. Possibilities for extensions of existing learning algorithms are discussed. Preliminary results are provided

关键词： machine learning machine intelligence Genetic algorithms Intelligent agent Neural networks Internet Biological neural networks Humans Learning systems Computational intelligence

来源：评论

学校读者我要写书评

暂无评论

Automatic image registration using virtual circles

引用

International Journal of Image and Graphics 2004年第2期4卷 281-299页

作者： Alhichri, Haikel S. Kamel, Mohamed Pattern Analysis and Machine Intelligence Laboratory Systems Design Engineering University of Waterloo 200 University Ave. West WaterlooONN2L3G1 Canada University of Waterloo Canada

The main contribution of this work is a novel set of image features called the virtual circles and their use in the registration of images under similarity transformations. A virtual circle is a circle with maximal radius encompassing a background area that does not contain edge points. It has many useful properties such as its radius, and its dominant edge direction for example, which can be utilized for efficient registration. Furthermore, virtual circles are frequent and can be extracted efficiently with the help of the distance transform from many types of images. We have tested the new virtual circles method in the registration of 66 pairs of images, half of which are printed labels and the other half are indoor scenes. Experimental results have shown that this method has a linear complexity in terms of the number of pixels. It is also highly automatic, because it has a small number of parameters, which almost never need to be changed throughout the experiments. © 2004 World Scientific Publishing Company.

关键词： Image registration

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：