Tractable Approximate Gaussian Inference for Bayesian Neural Networks
This addresses the computational bottleneck for researchers and practitioners in Bayesian deep learning, though it is incremental as it builds on existing approximate inference techniques.
The paper tackles the problem of intractable inference in Bayesian neural networks by proposing an analytical method for tractable approximate Gaussian inference (TAGI), which achieves computational complexity of O(n) and matches the performance of gradient-based methods on regression and classification benchmarks.
In this paper, we propose an analytical method for performing tractable approximate Gaussian inference (TAGI) in Bayesian neural networks. The method enables the analytical Gaussian inference of the posterior mean vector and diagonal covariance matrix for weights and biases. The method proposed has a computational complexity of $\mathcal{O}(n)$ with respect to the number of parameters $n$, and the tests performed on regression and classification benchmarks confirm that, for a same network architecture, it matches the performance of existing methods relying on gradient backpropagation.