Recurrent Attentive Neural Process for Sequential Data

TL;DR

This paper introduces RANP, a novel model combining attentive neural processes with recurrent neural networks to better capture temporal and sequential structures in data, outperforming previous models like NPs and LSTMs.

Contribution

The paper proposes RANP, integrating ANP with RNNs to enhance sequential data modeling and uncertainty estimation, addressing limitations of permutation invariance in neural processes.

Findings

RANP outperforms NPs and LSTMs in 1D regression tasks.

Effective modeling of sequential data and temporal order.

Improved uncertainty quantification in predictions.

Abstract

Neural processes (NPs) learn stochastic processes and predict the distribution of target output adaptively conditioned on a context set of observed input-output pairs. Furthermore, Attentive Neural Process (ANP) improved the prediction accuracy of NPs by incorporating attention mechanism among contexts and targets. In a number of real-world applications such as robotics, finance, speech, and biology, it is critical to learn the temporal order and recurrent structure from sequential data. However, the capability of NPs capturing these properties is limited due to its permutation invariance instinct. In this paper, we proposed the Recurrent Attentive Neural Process (RANP), or alternatively, Attentive Neural Process-RecurrentNeural Network(ANP-RNN), in which the ANP is incorporated into a recurrent neural network. The proposed model encapsulates both the inductive biases of recurrent neural networks and also the strength of NPs for modelling uncertainty. We demonstrate that RANP can effectively model sequential data and outperforms NPs and LSTMs remarkably in a 1D regression toy example as well as autonomous-driving applications.