本文研究概率神经符号学习中梯度计算的复杂度,提出基于模型采样的无偏梯度估计器 WeightME,并证明其在训练中可处理且仅需对数次 SAT 求解器调用,实验表明现有有偏近似在优化上存在困难。
摘要:
纯神经学习的局限性引发了人们对概率神经符号模型的兴趣,这类模型将神经网络与概率逻辑推理相结合。由于这些神经符号模型使用梯度下降进行训练,我们研究了概率推理求微分的复杂度。我们证明,尽管在一般情况下近似这些梯度是难以处理的,但在训练过程中它会变得可处理。此外,我们引入了 WeightME,一种基于模型采样的无偏梯度估计器。在温和假设下,WeightME 通过对 SAT 求解器的对数次调用,以概率保证近似梯度。最后,我们评估了这些保证在梯度上的必要性。我们的实验表明,即使在精确求解仍然可行的情况下,现有的有偏近似确实难以进行优化。
英文摘要:
The limitations of purely neural learning have sparked an interest in probabilistic neurosymbolic models, which combine neural networks with probabilistic logical reasoning. As these neurosymbolic models are trained with gradient descent, we study the complexity of differentiating probabilistic reasoning. We prove that although approximating these gradients is intractable in general, it becomes tractable during training. Furthermore, we introduce WeightME, an unbiased gradient estimator based on model sampling. Under mild assumptions, WeightME approximates the gradient with probabilistic guarantees using a logarithmic number of calls to a SAT solver. Lastly, we evaluate the necessity of these guarantees on the gradient. Our experiments indicate that the existing biased approximations indeed struggle to optimize even when exact solving is still feasible.
Department of Computer Science and Leuven.AI, KU Leuven, Belgium(1 计算机科学系和 Leuven.AI,KU Leuven,比利时);Centre for Applied Autonomous Sensor Systems (AASS), Örebro University, Sweden(3 应用自主传感器系统中心(AASS),Örebro大学,瑞典)