摘要:

A crucial dilemma is how to increase the power of connectionist networks (CN), since simply increasing the size of today's relatively small CNs often slows down and worsens learning and performance. There are three possible ways: (1) use more powerful structures; (2) increase the amount of stored information, and the power and the variety of the basic processes; (3) have the network modify itself (learn, evolve) in more powerful ways. Today's connectionist networks use only a few of the many possible topological structures, handle only numerical values using only very simple basic processes, and learn only by modifying weights associated with links. This paper examines the great variety of potentially muck more powerful possibilities, focusing on what appear to be the most promising: appropriate brain-like structures (e.g. local connectivity, global convergence and divergence); matching, symbol-handling, and list-manipulating capabilities; and learning by extraction-generation-discovery.

ISSN:0954-0091    

DOI:10.1080/09540099008915668

出版商:Taylor & Francis Group    

年代:1990

数据来源: Taylor

摘要:

We introduce a new connectionist paradigm which views neural networks as implementations of syntactic pattern recognition algorithms. Thus, learning is seen as a process of grammatical inference and recognition as a process of parsing. Naturally, the possible realizations of this theme are diverse; in this paper we present some initial explorations of the case where the pattern grammar is context-free, inferred (from examples) by a separate procedure, and then mapped onto a connectionist paper. Unlike most neural networks for which structure is pre-defined, the resulting network has as many levels as are necessary and arbitrary connections between levels. Furthermore, by the addition of a delay element, the network becomes capable of dealing with time-varying patterns in a simple and efficient manner. Since grammatical inference algorithms are notoriously expensive computationally, we place an important restriction on the type of context-free grammars which can be inferred. This dramatically reduces complexity. The resulting grammars are called ‘strictly-hierarchical’ and map straightforwardly onto a temporal connectionist parser (TCP) using a relatively small number of neurons. The new paradigm is applicable to a variety of pattern-processing tasks such as speech recognition and character recognition. We concentrate here on hand-written character recognition; performance in other problem domains will be reported in future publications. Results are presented to illustrate the performance of the system with respect to a number of parameters, namely, the inherent variability of the data, the nature of the learning (supervised or unsupervised) and the details of the clustering procedure used to limit the number of non-terminals inferred. In each of these cases (eight in total), we contrast the performance of a stochastic and a non-stochastic TCP. The stochastic TCP does have greater powers of discrimination, but in many cases the results were very similar. If this result holds in practical situations it is important, because the non-stochastic version has a straightforward implementation in silicon.

ISSN:0954-0091    

DOI:10.1080/09540099008915669

出版商:Taylor & Francis Group    

年代:1990

数据来源: Taylor

摘要:

A multi-layer neural network is used to control the navigation of a mobile robot in an environment containing obstacles in a temperature field. The mobile robot must avoid obstacles and hill climb towards the maximum of the temperature field. The strategy for each of these two tasks is acquired by learning. First by exploring the environment, the mobile robot extracts the relevant sensory situations by building up an internal map of the environment. The associations between these situations and the appropriate actions are then formed in an unsupervised manner, i.e. with no ‘teacher’required. The proposed structure of the system permits the coordination of the two tasks. Simulation results display not only the ability of the robot to achieve collision-free navigation towards its target in the explored environment, but also in new unvisited environments, illustrating the generalization property of neural networks.

ISSN:0954-0091    

DOI:10.1080/09540099008915670

出版商:Taylor & Francis Group    

年代:1990

数据来源: Taylor

摘要:

Text-to-speech systems usually consist of a preprocessor for expanding abbreviations, a system for converting orthographic text to a phonemic representation, rules for generating appropriate rhythm and intonation, and a speech synthesizer to generate an acoustic waveform from the phonemic representation. Multi-layer perceptrons have recently been used for the orthographic to phonemic conversion process. In this paper the possibility of using perceptrons in the preprocessor is explored. It is shown that single-layer perceptrons are sufficient for expanding 3-digit numbers, 4-digit numbers and cardinal numbers into appropriate orthographic text, but a multi-layer perceptron is required for expanding 12-hour clock times.

ISSN:0954-0091    

DOI:10.1080/09540099008915671

出版商:Taylor & Francis Group    

年代:1990

数据来源: Taylor

ISSN:0954-0091    

DOI:10.1080/09540099008915672

出版商:Taylor & Francis Group    

年代:1990

数据来源: Taylor

ISSN:0954-0091    

DOI:10.1080/09540099008915673

出版商:Taylor & Francis Group    

年代:1990

数据来源: Taylor

ISSN:0954-0091    

DOI:10.1080/09540099008928051

出版商:Taylor & Francis Group    

年代:1990

数据来源: Taylor