Multi-Task Neural Processes

TL;DR

Multi-Task Neural Processes (MTNPs) extend Neural Processes to jointly model multiple correlated tasks, including incomplete data scenarios, improving predictive performance on real-world multi-task datasets.

Contribution

MTNPs introduce a hierarchical model that captures inter-task correlations and handles incomplete data, advancing multi-task inference in stochastic process modeling.

Findings

Successfully model multiple correlated tasks in real-world data

Exploit inter-task correlations to improve predictions

Handle incomplete multi-task data effectively

Abstract

Neural Processes (NPs) consider a task as a function realized from a stochastic process and flexibly adapt to unseen tasks through inference on functions. However, naive NPs can model data from only a single stochastic process and are designed to infer each task independently. Since many real-world data represent a set of correlated tasks from multiple sources (e.g., multiple attributes and multi-sensor data), it is beneficial to infer them jointly and exploit the underlying correlation to improve the predictive performance. To this end, we propose Multi-Task Neural Processes (MTNPs), an extension of NPs designed to jointly infer tasks realized from multiple stochastic processes. We build MTNPs in a hierarchical way such that inter-task correlation is considered by conditioning all per-task latent variables on a single global latent variable. In addition, we further design our MTNPs so that they can address multi-task settings with incomplete data (i.e., not all tasks share the same set of input points), which has high practical demands in various applications. Experiments demonstrate that MTNPs can successfully model multiple tasks jointly by discovering and exploiting their correlations in various real-world data such as time series of weather attributes and pixel-aligned visual modalities. We release our code at https://github.com/GitGyun/multi_task_neural_processes.