Equivariant Learning of Stochastic Fields: Gaussian Processes and Steerable Conditional Neural Processes
arXiv:2011.12916
Abstract
Motivated by objects such as electric fields or fluid streams, we study the problem of learning stochastic fields, i.e. stochastic processes whose samples are fields like those occurring in physics and engineering. Considering general transformations such as rotations and reflections, we show that spatial invariance of stochastic fields requires an inference model to be equivariant. Leveraging recent advances from the equivariance literature, we study equivariance in two classes of models. Firstly, we fully characterise equivariant Gaussian processes. Secondly, we introduce Steerable Conditional Neural Processes (SteerCNPs), a new, fully equivariant member of the Neural Process family. In experiments with Gaussian process vector fields, images, and real-world weather data, we observe that SteerCNPs significantly improve the performance of previous models and equivariance leads to improvements in transfer learning tasks.
References in corpus (17)
- The NumPy array: a structure for efficient numerical computation
- Group Equivariant Convolutional Networks
- On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups
- 3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data
- Spherical CNNs
- E(n) Equivariant Graph Neural Networks
- General -Equivariant Steerable CNNs
- Incorporating Symmetry into Deep Dynamics Models for Improved Generalization
- Conditional Neural Processes
- Neural Processes
- LieTransformer: Equivariant self-attention for Lie Groups
- Harmonic Networks: Deep Translation and Rotation Equivariance
- Meta-Learning Symmetries by Reparameterization
- Meta-Learning Stationary Stochastic Process Prediction with Convolutional Neural Processes
- Deep Kernel Learning
- Consistent Generative Query Networks
- The Gaussian Neural Process