Learning Algorithms via Neural Logic Networks
This addresses the need for interpretable and explicit learning in deep neural networks for researchers and practitioners in AI, particularly in inductive logic programming, though it appears incremental as it builds on existing neural ILP methods.
The paper tackles the problem of learning Boolean functions and discrete-algorithmic tasks by proposing Neural Logic Networks (NLNs), a novel learning paradigm using Boolean logic algebra, and demonstrates that NLNs can outperform current state-of-the-art neural ILP solvers on benchmarks like decimal addition, multiplication, and sorting.
We propose a novel learning paradigm for Deep Neural Networks (DNN) by using Boolean logic algebra. We first present the basic differentiable operators of a Boolean system such as conjunction, disjunction and exclusive-OR and show how these elementary operators can be combined in a simple and meaningful way to form Neural Logic Networks (NLNs). We examine the effectiveness of the proposed NLN framework in learning Boolean functions and discrete-algorithmic tasks. We demonstrate that, in contrast to the implicit learning in MLP approach, the proposed neural logic networks can learn the logical functions explicitly that can be verified and interpreted by human. In particular, we propose a new framework for learning the inductive logic programming (ILP) problems by exploiting the explicit representational power of NLN. We show the proposed neural ILP solver is capable of feats such as predicate invention and recursion and can outperform the current state of the art neural ILP solvers using a variety of benchmark tasks such as decimal addition and multiplication, and sorting on ordered list.