Learning Interpretable Differentiable Logic Networks
This addresses interpretability and efficiency issues in neural networks for applications like healthcare and NLP, though it is an incremental improvement by adapting existing logic-based methods with gradient-based learning.
The paper tackles the problem of neural networks being black-box and prone to overfitting by introducing differentiable logic networks (DLNs) that use binary logic operators, achieving accuracies comparable to or exceeding traditional NNs on twenty classification tasks and reducing inference operations by up to a thousand times.
The ubiquity of neural networks (NNs) in real-world applications, from healthcare to natural language processing, underscores their immense utility in capturing complex relationships within high-dimensional data. However, NNs come with notable disadvantages, such as their "black-box" nature, which hampers interpretability, as well as their tendency to overfit the training data. We introduce a novel method for learning interpretable differentiable logic networks (DLNs) that are architectures that employ multiple layers of binary logic operators. We train these networks by softening and differentiating their discrete components, e.g., through binarization of inputs, binary logic operations, and connections between neurons. This approach enables the use of gradient-based learning methods. Experimental results on twenty classification tasks indicate that differentiable logic networks can achieve accuracies comparable to or exceeding that of traditional NNs. Equally importantly, these networks offer the advantage of interpretability. Moreover, their relatively simple structure results in the number of logic gate-level operations during inference being up to a thousand times smaller than NNs, making them suitable for deployment on edge devices.