Saad Emad W, Wunsch Donald C
Phantom Works, The Boeing Company, Seattle, WA 98124, United States.
Neural Netw. 2007 Jan;20(1):78-93. doi: 10.1016/j.neunet.2006.07.005. Epub 2006 Oct 6.
An important drawback of many artificial neural networks (ANN) is their lack of explanation capability [Andrews, R., Diederich, J., & Tickle, A. B. (1996). A survey and critique of techniques for extracting rules from trained artificial neural networks. Knowledge-Based Systems, 8, 373-389]. This paper starts with a survey of algorithms which attempt to explain the ANN output. We then present HYPINV, a new explanation algorithm which relies on network inversion; i.e. calculating the ANN input which produces a desired output. HYPINV is a pedagogical algorithm, that extracts rules, in the form of hyperplanes. It is able to generate rules with arbitrarily desired fidelity, maintaining a fidelity-complexity tradeoff. To our knowledge, HYPINV is the only pedagogical rule extraction method, which extracts hyperplane rules from continuous or binary attribute neural networks. Different network inversion techniques, involving gradient descent as well as an evolutionary algorithm, are presented. An information theoretic treatment of rule extraction is presented. HYPINV is applied to example synthetic problems, to a real aerospace problem, and compared with similar algorithms using benchmark problems.
许多人工神经网络(ANN)的一个重要缺点是它们缺乏解释能力[安德鲁斯,R.,迪德里希,J.,& 蒂克尔,A. B.(1996年)。对从训练好的人工神经网络中提取规则的技术的调查与批判。基于知识的系统,8,373 - 389]。本文首先对试图解释ANN输出的算法进行了调查。然后我们提出了HYPINV,一种新的解释算法,它依赖于网络反演;即计算产生期望输出的ANN输入。HYPINV是一种教学算法,它以超平面的形式提取规则。它能够以任意期望的保真度生成规则,保持保真度与复杂度的权衡。据我们所知,HYPINV是唯一一种从连续或二元属性神经网络中提取超平面规则的教学规则提取方法。文中介绍了不同的网络反演技术,包括梯度下降以及一种进化算法。还给出了规则提取的信息论处理方法。HYPINV被应用于示例合成问题、一个实际的航空航天问题,并与使用基准问题的类似算法进行了比较。