This article presents a novel method, called Modular Grammatical Evolution (MGE), toward validating the hypothesis that restricting the solution space of NeuroEvolution to modular and simple neural networks enables the efficient generation of smaller and more structured neural networks while providing acceptable (and in some cases superior) accuracy on large data sets. MGE also enhances the state-of-the-art Grammatical Evolution (GE) methods in two directions. First, MGE's representation is modular in that each individual has a set of genes, and each gene is mapped to a neuron by grammatical rules. Second, the proposed representation mitigates two important drawbacks of GE, namely the low scalability and weak locality of representation, toward generating modular and multilayer networks with a high number of neurons. We define and evaluate five different forms of structures with and without modularity using MGE and find single-layer modules with no coupling more productive. Our experiments demonstrate that modularity helps in finding better neural networks faster. We have validated the proposed method using ten well-known classification benchmarks with different sizes, feature counts, and output class counts. Our experimental results indicate that MGE provides superior accuracy with respect to existing NeuroEvolution methods and returns classifiers that are significantly simpler than other machine learning generated classifiers. Finally, we empirically demonstrate that MGE outperforms other GE methods in terms of locality and scalability properties.

本文提出了一种称为模块化语法进化 (MGE) 的新方法，以验证将 NeuroEvolution 的解决方案空间限制为模块化和简单的神经网络能够有效生成更小、更结构化的神经网络，同时提供可接受的（在某些情况下）案例优越）在大型数据集上的准确性。MGE 还在两个方向上增强了最先进的语法进化 (GE) 方法。首先，MGE 的表示是模块化的，因为每个个体都有一组基因，每个基因通过语法规则映射到一个神经元。其次，所提出的表示减轻了 GE 的两个重要缺点，即表示的低可扩展性和弱局部性，以生成具有大量神经元的模块化和多层网络。我们使用 MGE 定义和评估具有和不具有模块化的五种不同形式的结构，并发现没有耦合的单层模块更有效率。我们的实验表明，模块化有助于更快地找到更好的神经网络。我们已经使用十个具有不同大小、特征计数和输出类计数的知名分类基准验证了所提出的方法。我们的实验结果表明，相对于现有的 NeuroEvolution 方法，MGE 提供了更高的准确性，并且返回的分类器比其他机器学习生成的分类器要简单得多。最后，我们凭经验证明 MGE 在局部性和可扩展性属性方面优于其他 GE 方法。