Efficient Modeling of Latent Information in Supervised Learning using Gaussian Processes
This work addresses the challenge of generalizing to new conditions with limited data in machine learning, which is incremental as it builds on existing Gaussian process methods.
The paper tackles the problem of modeling latent information across multiple conditions in supervised learning, such as voice recordings from different persons, and introduces Latent Variable Multiple Output Gaussian Processes (LVMOGP) to generalize to new conditions with few data points, showing significant outperformance over related Gaussian process methods on synthetic and real data.
Often in machine learning, data are collected as a combination of multiple conditions, e.g., the voice recordings of multiple persons, each labeled with an ID. How could we build a model that captures the latent information related to these conditions and generalize to a new one with few data? We present a new model called Latent Variable Multiple Output Gaussian Processes (LVMOGP) and that allows to jointly model multiple conditions for regression and generalize to a new condition with a few data points at test time. LVMOGP infers the posteriors of Gaussian processes together with a latent space representing the information about different conditions. We derive an efficient variational inference method for LVMOGP, of which the computational complexity is as low as sparse Gaussian processes. We show that LVMOGP significantly outperforms related Gaussian process methods on various tasks with both synthetic and real data.