arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1909.05862cs.LGastro-ph.IMphysics.comp-phstat.ML
收藏

用图网络学习符号物理

Learning Symbolic Physics with Graph Networks

Princeton University(普林斯顿大学);DeepMind(深度思考);Flatiron Institute(弗莱提荣研究所)

Miles D. Cranmer, Rui Xu, Peter Battaglia, Shirley Ho

摘要:

我们介绍了一种方法,用于在图网络上施加物理动机的归纳偏差,以学习可解释的表示和改进的zero-shot泛化。我们的实验表明,我们的图网络模型实现了这种归纳偏差,当在n体引力和弹簧模拟上训练时,可以学习相当于真实力矢量的消息表示。我们使用符号回归将显式代数方程拟合到训练模型的消息函数,并在没有先验知识的情况下恢复牛顿引力定律的符号形式。我们还表明,我们的模型在推理时间比在训练过程中经历的身体更多的系统更好地推广。原则上,我们的方法可扩展到由图网络学习的任何未知的相互作用规律,并提供了一种有价值的技术,用于从深度学习捕获的隐式知识中解释和推断关于世界的显式因果理论。

英文摘要:

We introduce an approach for imposing physically motivated inductive biases on graph networks to learn interpretable representations and improved zero-shot generalization. Our experiments show that our graph network models, which implement this inductive bias, can learn message representations equivalent to the true force vector when trained on n-body gravitational and spring-like simulations. We use symbolic regression to fit explicit algebraic equations to our trained model's message function and recover the symbolic form of Newton's law of gravitation without prior knowledge. We also show that our model generalizes better at inference time to systems with more bodies than had been experienced during training. Our approach is extensible, in principle, to any unknown interaction law learned by a graph network, and offers a valuable technique for interpreting and inferring explicit causal theories about the world from implicit knowledge captured by deep learning.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑