Bayesian Nonlinear Support Vector Machines for Big Data
This work addresses the challenge of efficient Bayesian inference for large-scale nonlinear classification, offering improvements in speed and scalability for practitioners in machine learning.
The authors tackled the problem of scaling Bayesian nonlinear support vector machines to big data by proposing a fast inference method using stochastic variational inference and inducing points, resulting in a method that is faster than competing Bayesian approaches and scales to millions of data points.
We propose a fast inference method for Bayesian nonlinear support vector machines that leverages stochastic variational inference and inducing points. Our experiments show that the proposed method is faster than competing Bayesian approaches and scales easily to millions of data points. It provides additional features over frequentist competitors such as accurate predictive uncertainty estimates and automatic hyperparameter search.