Scalable Gaussian Process Variational Autoencoders
This work addresses scalability issues in GP-VAEs for researchers and practitioners in machine learning, representing an incremental improvement over prior methods.
The authors tackled the scalability limitations of Gaussian Process Variational Autoencoders (GP-VAEs) by introducing a new model that uses sparse inference approaches, resulting in improved runtime and memory efficiency compared to existing methods.
Conventional variational autoencoders fail in modeling correlations between data points due to their use of factorized priors. Amortized Gaussian process inference through GP-VAEs has led to significant improvements in this regard, but is still inhibited by the intrinsic complexity of exact GP inference. We improve the scalability of these methods through principled sparse inference approaches. We propose a new scalable GP-VAE model that outperforms existing approaches in terms of runtime and memory footprint, is easy to implement, and allows for joint end-to-end optimization of all components.