神经网络的知识获取与行为解释
Knowledge Acquisition and Behavioral Explanation on Neural Networks
-
摘要: 神经网络的缺点之一在于它无法明显地表达其行为的含义.针对神经网络的不同结构, 通过给出节点取值不同的约束条件,描述了如何从神经网络提取知识,并对神经网络的行为进 行解释.举例说明了从神经网络提取产生式规划的策略.Abstract: One of the drawbacks of neural nets is that they can not explicitly express the meaning of their behaviors. This paper describes some problems on knowledge acquisition and behavioral explanation for neural nets with different structures, the constraints conditions are the numbers of nodes, and suggest a strategy of generating production rules for neural nets as an example.
-
Key words:
- neural net /
- knowledge acquisition /
- behavioral explanation /
- production rule
计量
- 文章访问数: 2380
- HTML全文浏览量: 175
- PDF下载量: 1095
- 被引次数: 0