Predictive Precompute with Recurrent Neural Networks

TL;DR

This paper introduces the use of recurrent neural networks to improve predictive precompute in applications, leading to more accurate predictions, less feature engineering, and significantly reduced computational costs at scale.

Contribution

It presents a novel application of RNNs for predictive precompute, demonstrating advantages over traditional models in accuracy and efficiency in large-scale deployment.

Findings

RNNs outperform traditional models in prediction accuracy.

RNNs reduce feature engineering requirements.

Prediction serving costs decrease by an order of magnitude.

Abstract

In both mobile and web applications, speeding up user interface response times can often lead to significant improvements in user engagement. A common technique to improve responsiveness is to precompute data ahead of time for specific activities. However, simply precomputing data for all user and activity combinations is prohibitive at scale due to both network constraints and server-side computational costs. It is therefore important to accurately predict per-user application usage in order to minimize wasted precomputation ("predictive precompute"). In this paper, we describe the novel application of recurrent neural networks (RNNs) for predictive precompute. We compare their performance with traditional machine learning models, and share findings from their large-scale production use at Facebook. We demonstrate that RNN models improve prediction accuracy, eliminate most feature engineering steps, and reduce the computational cost of serving predictions by an order of magnitude.