作者：
Hiroaki NakamuraIBM Research
Tokyo Research Laboratory 1623-14 Shimotsuruma Yamato-shi Kanagawa-ken 242-8502 Japan
The need for incremental algorithms for evaluating database queries is well known, but constructing algorithms that work on object-oriented databases (OODBs) has been difficult. The reason is that OODB query languages...
详细信息
ISBN:
(纸本)1581133359
The need for incremental algorithms for evaluating database queries is well known, but constructing algorithms that work on object-oriented databases (OODBs) has been difficult. The reason is that OODB query languages involve complex data types including composite objects and nested collections. As a result, existing algorithms have limitations in that the kinds of database updates are restricted, the operations found in many query languages are not supported, or the algorithms are too complex to be described precisely. We present an incremental computation algorithm that can handle any kind of database updates, can accept any expressions in complex query languages such as OQL, and can be described precisely. By translating primitive values and records into collections, we can reduce all query expressions comprehension. This makes the problems with incremental computation less complicated and thus allows us to decribe of two parts: one is to maintain the consistency in each comprehension occurrence and the other is to update the value of an entire expression. The algorithm is so flexible that we can use strict updates, lazy updates, and their combinations. By comparing the performance of applications built with our mechanism and that of equivalent hand written update programs, we show that our incremental algorithm can be iplemented efficiently.
The efficacy of a novel fuzzy neural network classifier for the characterization of ultrasonic liver images based on texture analysis techniques is investigated. Classification features are extracted with the use of i...
详细信息
The efficacy of a novel fuzzy neural network classifier for the characterization of ultrasonic liver images based on texture analysis techniques is investigated. Classification features are extracted with the use of image texture analysis techniques such as fractal dimension texture analysis, spatial gray-level dependence matrices, gray-level difference statistics, gray-level run-length statistics, and first-order gray-level parameters. These features are fed to a neural network classifier based on geometrical fuzzy sets. Starting from the construction of the Voronoi diagram of the training patterns, an aggregation of Voronoi regions is performed, leading to the identification of larger regions belonging exclusively to one of the pattern classes. The resulting scheme is a constructive algorithm that defines fuzzy clusters of patterns. Based on observations concerning the grade of membership of the training patterns to the created regions, decision probabilities are computed through which the final classification is performed.
This article proves that the task of computing near-optimal weights for sigmoidal nodes under the L-1 regression norm is NP-Hard. For the special case where the sigmoid is piecewise linear, we prove a slightly stronge...
详细信息
This article proves that the task of computing near-optimal weights for sigmoidal nodes under the L-1 regression norm is NP-Hard. For the special case where the sigmoid is piecewise linear, we prove a slightly stronger result: that computing the optimal weights is NP-Hard. These results parallel that for the one-node pattern recognition problem-that determining the optimal weights for a threshold logic node is also intractable. Our results have important consequences for constructive algorithms that build a regression model one node at a time. It suggests that although such methods are (in principle) capable of producing efficient size representations (Barron, 1993;Jones, 1992), finding such representations may be computationally intractable. These results holds only in the deterministic sense;that is, they do not exclude the possibility that such representations may be found efficiently with high probability. In fact it motivates the use of heuristic or randomized algorithms for this problem.
A new constructive algorithm is presented for building neural networks that learn to reproduce output temporal sequences based on one or several input sequences. This algorithm builds a network for the task of system ...
详细信息
A new constructive algorithm is presented for building neural networks that learn to reproduce output temporal sequences based on one or several input sequences. This algorithm builds a network for the task of system modelling, dealing with continuous variables in the discrete rime domain. The constructive scheme makes it user independent. The network's structure consists of an ordinary set and a classification set, so it is a hybrid network like that of Stokbro et al. [6], but with a binary classification. The networks can easily be interpreted, so the learned representation can be transferred to a human engineer, unlike many other network models. This allows for a better understanding of the system structure than just its simulation. This constructive algorithm limits the network complexity automatically, hence preserving extrapolation capabilities. Examples with real data from three totally different sources show good performance and allow for a promising line of research.
In this survey paper, we review the constructive algorithms for structure learning in feedforward neural networks for regression problems, The basic idea is to start with a small network, then add hidden units and wei...
详细信息
In this survey paper, we review the constructive algorithms for structure learning in feedforward neural networks for regression problems, The basic idea is to start with a small network, then add hidden units and weights incrementally until a satisfactory solution is found, By formulating the whole problem as a state-space search, we first describe the general issues in constructive algorithms, with special emphasis on the search strategy, A taxonomy, based on the differences in the state transition mapping, the training algorithm, and the network architecture, is then presented.
We consider a spatially uniform asymptotic representation at large times of the solution to the Cauchy problem for the non-linear Schrodinger equation. If the non-linear term decreases in time faster than the linear t...
详细信息
We consider a spatially uniform asymptotic representation at large times of the solution to the Cauchy problem for the non-linear Schrodinger equation. If the non-linear term decreases in time faster than the linear terms, then the asymptotics are quasi-linear. Of particular interest is the case in which the non-linearity decreases in time at the same rate as or even more slowly than the linear terms and thus has a stronger effect on the solution asymptotics at large times. In this paper we employ an appropriate change of variables to reduce this case to the quasi-linear one. Namely, we derive an integral equation with rapidly decreasing non-linearity for the new unknown function, which can be solved by the method of successive approximations. Thus, we have a constructive algorithm for calculating the asymptotics of the solution to the Cauchy problem for the non-linear Schrodinger equation from the initial data.
A new constructive algorithm for designing and training multilayer perceptrons is proposed. This algorithm involves the optimization of an objective function for internal representations, which does not require any co...
详细信息
A new constructive algorithm for designing and training multilayer perceptrons is proposed. This algorithm involves the optimization of an objective function for internal representations, which does not require any computation of the network's outputs. Coupled with a strategy for recruiting units during the learning process, this concept provides a scheme for training a multilayer network layer by layer, until self-encoding of the pattern categories is achieved in the final, highest-level representations. Two objective functions are proposed For discrimination problems, recent experimental and theoretical results concerning back-propagation training of networks with one hidden layer and linear outputs suggest the introduction of a particular measure of class separability. For problems involving the approximation of a continuous function, we show that the minimization of the mean squared output error can be achieved by maximizing a statistical measure (the sample coefficient of multiple determination) in the last hidden layer. Simulations are used to illustrate the process of network construction, and to demonstrate the improvements brought by this approach over back-propagation in terms of performance and robustness.
Feature construction has been shown to be an useful technique to improve the efficiency of extracting information from raw data. We develop a modified feature construction algorithm, using correlation information amon...
详细信息
Feature construction has been shown to be an useful technique to improve the efficiency of extracting information from raw data. We develop a modified feature construction algorithm, using correlation information among the initial set of features, and study its performance. Feed-forward neural networks, using the back-propagation algorithm to learn, have been shown to have excellent properties while plagued with the problem of time needed to learn concepts. We alleviate this inherent problem with the back-propagation algorithm using data pre-processed by the proposed feature construction algorithm. Initial results suggest that this methodology can be beneficially used along with other means of improving the learning performance in feed-forward neural networks.
This paper extends the classical notion of critical paths in combinational circuits to the case of synchronous circuits that use level-sensitive latches. Critical paths in such circuits arise from setup, hold, and cyc...
详细信息
This paper extends the classical notion of critical paths in combinational circuits to the case of synchronous circuits that use level-sensitive latches. Critical paths in such circuits arise from setup, hold, and cyclic constraints on the data signals at the inputs of each latch and may extend through one or more latches. Two approaches are presented for identifying these critical paths and verifying their timing. The first implicitly checks all paths using a relaxation-based solution procedure. Results of this procedure are used to calculate slack values, which in turn identify satisfied and violated critical paths. The second approach is based on a constructive algorithm which generates all the critical paths in a circuit and then verifies that their timing constraints are satisfied. algorithms are evaluated and compared using circuits from the ISCAS89 sequential benchmark suite and the Michigan High Performance Microprocessor Project.
The analysis of an example illustrates the high "resolution" of group analysis of the system (1), (2); only three of the nine symmetry operators found can be determined by means of the -theorem. With the aid...
详细信息
The analysis of an example illustrates the high "resolution" of group analysis of the system (1), (2); only three of the nine symmetry operators found can be determined by means of the -theorem. With the aid of the above-determined symmetry operators it can be confirmed that two essential parameters can be formed out of the eight parameters of the system (1), (2): the thrust—weight ratio (P) and the lift—drag ratio (k). The application of group methods leads to constructive algorithms for seeking the required transformations, which the investigator can use to find the substitutions needed for any problem at hand.
暂无评论