The non-linear dynamics of a chaotic attractor offer a number of useful features to the developer of neuromorphic systems. Included in these is the ability for efficient memory storage and recall. A chaotic attractor ...
详细信息
The non-linear dynamics of a chaotic attractor offer a number of useful features to the developer of neuromorphic systems. Included in these is the ability for efficient memory storage and recall. A chaotic attractor has a potentially infinite number of Unstable Periodic Orbits (UPO) embedded within it. These orbits can be stabilised with the application of delayed feedback inhibition. This research investigates the possibility of using such delayed feedback in a network to stabiles different UPOs in response to disparate input stimuli. A key feature of the models presented is that the UPOs, which correspond to dynamic memory states, emerge from the dynamics of the attractor. The paper presents two learning rules which support the network dynamics from which the memory states emerge.
This paper proposes a Rectangle Expansion Model, which is deduced from active contour model, especially for page layout analysis. The advantage of this model is the robust flexibility no needing to set so many thresho...
详细信息
ISBN:
(纸本)0780375084
This paper proposes a Rectangle Expansion Model, which is deduced from active contour model, especially for page layout analysis. The advantage of this model is the robust flexibility no needing to set so many thresholds as conventional method. Page Layout Analysis is an important part for OCR system to implement printed documents digitization, so this approach can improve the performance of character recognition.
Support Vector Machines (SVMs) have played a key role in broad classes of problems in various fields. However, with increasing amounts of data being generated by businesses and researchers, SVMs suffer from the proble...
详细信息
ISBN:
(纸本)0780375084
Support Vector Machines (SVMs) have played a key role in broad classes of problems in various fields. However, with increasing amounts of data being generated by businesses and researchers, SVMs suffer from the problem of large memory requirement and CPU time when trained in batch mode on large data set The training process involves the solution of a quadratic programming problem. In this paper, we attempt to overcome these limitations and propose an approach based on incremental learning technique and multiple proximal support vector machines classifier. Experiment on generated data set gives promising result.
Performance analyses of in building distributed antenna (DA) systems for W-CDMA are presented. Uplink and downlink outage probabilities are related to the transmitted power, the shadowing statistics and the loading of...
详细信息
Performance analyses of in building distributed antenna (DA) systems for W-CDMA are presented. Uplink and downlink outage probabilities are related to the transmitted power, the shadowing statistics and the loading of the cell and it is observed that by deploying DA's in a multi-floor environment the advantages in terms of both coverage and capacity are significant. However, capacity limitations are present due to the noise rise of home cell power controlled interferers, which give rise over thermal noise. In order to overcome these limitations two more flexible architectures, namely zoning and switching, are presented and compared.
This paper describes an approach to the provision of a Structured Spreadsheet Engineering Methodology. The proposed methodology is mainly based on the classical systems development life cycle, structured methods and s...
详细信息
This paper describes an approach to the provision of a Structured Spreadsheet Engineering Methodology. The proposed methodology is mainly based on the classical systems development life cycle, structured methods and software engineering principles. It addresses the widespread problem of spreadsheet errors and is an extension to published work by Chadwick-97, Rajalingham-98, Rajalingham-99, Rajalingham-99a, Rajalingham-00, Rajalingham-00a, Rajalingham-00b and Rajalingham-01. This methodology also helps in training users in the process of spreadsheet building. Although there are variations of the life cycle for systems development, they are fundamentally similar to each other. The proposed Structured Spreadsheet Engineering Methodology is primarily based on the systems development life cycle described by Aktas-85, Jackson structures (Jackson-75) and approaches recommended by other authors. Numerous approaches are incorporated into this framework, making it a highly integrated and structured methodology for spreadsheet design and development. Apart from the concepts and principles borrowed from the above methods, the methodology also contains new developments in the research into integrity control of spreadsheet models.
Dynamic modeling for flexible manipulators is studied in this article. At first, we introduce a new dynamic modeling approach, and discuss the representations of the system general properties with screw theory. Then, ...
详细信息
Dynamic modeling for flexible manipulators is studied in this article. At first, we introduce a new dynamic modeling approach, and discuss the representations of the system general properties with screw theory. Then, we try to illustrate the effectiveness of the new modeling approach by comparative modeling and analysis with the finite element method. At last, we convince the modeling accuracy of the Ding-Holzer method with the results of concrete numerical simulations. Our study shows that the Ding-Holzer method is very simple and accurate to the dynamic modeling of flexible manipulators.
This paper presents a new method of constructing multi-class SVM classifier, which is based on the structure of Decision Directed Acyclic Graph (DDAG) and using active constraint for each SVM classifier. For k-class p...
详细信息
ISBN:
(纸本)0780375084
This paper presents a new method of constructing multi-class SVM classifier, which is based on the structure of Decision Directed Acyclic Graph (DDAG) and using active constraint for each SVM classifier. For k-class problem, it combines k(k-1)/2 two-class SVM classifiers, one for each pair of classes. In order to speed up the training and decision process of the classifier, we make three changes on the standard two-class classifiers, i.e. large margin, 2-norm squared for the error for the soft margin and active constraint. While in the testing phase, we use a rooted binary directed acyclic graph which has k(k-1)/2 internal nodes and k leaves. Computational experiment indicates that this is a simple and fast approach to generate multi-class SVM classifiers.
We study the problem of computing classification rule sets from relational databases so that accurate predictions can be made on test data with missing attribute values. Traditional classifiers perform badly when test...
详细信息
ISBN:
(纸本)9781581135671
We study the problem of computing classification rule sets from relational databases so that accurate predictions can be made on test data with missing attribute values. Traditional classifiers perform badly when test data are not as complete as the training data because they tailor a training database too much. We introduce the concept of one rule set being more robust than another, that is, able to make more accurate predictions on test data with missing attribute values. We show that the optimal class association rule set is as robust as the complete class association rule set. We then introduce the k-optimal rule set, which provides predictions exactly the same as the optimal class association rule set on test data with up to k missing attribute values. This leads to a hierarchy of k-optimal rule sets in which decreasing size corresponds to decreasing robustness, and they all more robust than a traditional classification rule set. We introduce two methods to find k-optimal rule sets, i.e. an optimal association rule mining approach and a heuristic approximate approach. We show experimentally that a k-optimal rule set generated by the optimal association rule mining approach performs better than that by the heuristic approximate approach and both rule sets perform significantly better than a typical classification rule set (C4.5Rules) on incomplete test data.
The analysis of the dynamical response of gas filled cavities surrounded by elastic capsules, acting as contrast agent microbubbles, is of importance in elucidating their full diagnostic use for ultrasound imaging. Wh...
详细信息
The analysis of the dynamical response of gas filled cavities surrounded by elastic capsules, acting as contrast agent microbubbles, is of importance in elucidating their full diagnostic use for ultrasound imaging. When the frequency of the external acoustic field is at or near the natural frequency of the bubble, resulting in resonance, larger amplitudes of oscillation than would otherwise occur are observed, enhancing the backscattered signal. Hence, it is important to medical ultrasound imaging to be able to accurately determine the complex resonant behavior of the bubble motion.
暂无评论