|
1. |
(S)RAAM: An Analytical Technique for Fast and Reliable Derivation of Connectionist Symbol Structure Representations |
|
Connection Science,
Volume 9,
Issue 2,
1997,
Page 139-160
ROBERT E CALLAN,
DOMINIC PALMER-BROWN,
Preview
|
PDF (290KB)
|
|
摘要:
Recursive auto-associative memory (RAAM) has become established in the connectionist literature as a key contribution in the strive to develop connectionist representations of symbol structures. However, RAAMs use the backpropagation algorithm and therefore can be difficult to train and slow to learn. In addition, it is often hard to analyze exactly what a network has learnt and, therefore, it is difficult to state what composition mechanism is used by a RAAM for constructing representations. In this paper, we present an analytical version of RAAM, denoted as simplified RAAM or (S)RAAM. (S)RAAM models a RAAM very closely in that a single constructor matrix is derived which can be applied recursively to construct connectionist representations of symbol structures. The derivation, like RAAM, exhibits a moving target effect because training patterns adjust during learning but, unlike RAAM, the training is very fast. The analytical model allows a clear statement to be made about generalization characteristics and it can be shown that, in practice, the model will converge.
ISSN:0954-0091
DOI:10.1080/095400997116667
出版商:Taylor & Francis Group
年代:1997
数据来源: Taylor
|
2. |
Visual Schemas in Neural Networks for Object Recognition and Scene Analysis |
|
Connection Science,
Volume 9,
Issue 2,
1997,
Page 161-200
WEE KHENG LEOW,
RISTO MIIKKULAINEN,
Preview
|
PDF (883KB)
|
|
摘要:
VISOR is a large connectionist system that shows how visual schemas can be learned, represented and used through mechanisms natural to neural networks. Processing in VISOR is based on cooperation, competition, and parallel bottom-up and top-down activation of schema representations. VISOR is robust against noise and variations in the inputs and parameters. It can indicate the confidence of its analysis, pay attention to important minor differences, and use context to recognize ambiguous objects. Experiments also suggest that the representation and learning are stable, and behavior is consistent with human processes such as priming, perceptual reversal and circular reaction in learning. The schema mechanisms of VISOR can serve as a starting point for building robust high-level vision systems, and perhaps for schema-based motor control and natural language processing systems as well.
ISSN:0954-0091
DOI:10.1080/095400997116676
出版商:Taylor & Francis Group
年代:1997
数据来源: Taylor
|
3. |
Complex Connectionist Dynamics and the Limited Complexity of the Resulting Functions |
|
Connection Science,
Volume 9,
Issue 2,
1997,
Page 201-216
ACHIM G HOFFMANN,
Preview
|
PDF (241KB)
|
|
摘要:
The paper demonstrates how algorithmic information theory can be elegantly used as a powerful tool for analyzing the dynamics in connectionist systems. It is shown that simple structures of connectionist systems-even if they are very large-are unable significantly to ease the problem of learning complex functions. Also, the development of new learning algorithms would not essentially change this situation. Lower and upper bounds are given for the number of examples needed to learn complex concepts. The bounds are proved with respect to the notion of probably approximately correct learning. It is proposed to use algorithmic information theory for further studies on network dynamics.
ISSN:0954-0091
DOI:10.1080/095400997116685
出版商:Taylor & Francis Group
年代:1997
数据来源: Taylor
|
4. |
Coding the Output of a Feedforward Neural Net |
|
Connection Science,
Volume 9,
Issue 2,
1997,
Page 217-228
KHALID A AL-MASHOUQ,
Preview
|
PDF (189KB)
|
|
摘要:
In any feedforward neural net, there are many choices for coding, or labeling, the input data. Error-correcting codes have been proposed to encode the neural net output. This representation adds extra distance between the labels of the different classes, thus it helps to combat classification errors encountered in feedforward nets. This claim is verified theoretically and some useful bounds are derived to characterize the error-correcting code for such applications. The overhead of coding is to require more output nodes, i.e. a more complex network. It is known that, in general a more complex network has more capacity. Thus, we investigate the capacity and separation ability of the coded network and compare the effect of coding the output with that of using a specific two-layered net. This comparison is carried out from a deterministic and then from a probabilistic view point. The issues of finding a neural net decoder is also addressed and analyzed. This leads to a new look at the multi-layer neural net which helps in finding an upper bound on the complexity of the multi-layer neural net.
ISSN:0954-0091
DOI:10.1080/095400997116694
出版商:Taylor & Francis Group
年代:1997
数据来源: Taylor
|
|