Feature Staleness Aware Incremental Learning for CTR Prediction

TL;DR

This paper introduces FeSAIL, a novel incremental learning method for CTR prediction that adaptively replays stale feature samples to prevent performance degradation, outperforming existing methods on benchmark datasets.

Contribution

FeSAIL is the first approach to address feature staleness in incremental CTR learning by combining staleness aware sampling and regularization mechanisms.

Findings

FeSAIL achieves superior accuracy on four benchmark datasets.

FeSAIL effectively mitigates feature staleness issues.

The method improves training efficiency and model robustness.

Abstract

Click-through Rate (CTR) prediction in real-world recommender systems often deals with billions of user interactions every day. To improve the training efficiency, it is common to update the CTR prediction model incrementally using the new incremental data and a subset of historical data. However, the feature embeddings of a CTR prediction model often get stale when the corresponding features do not appear in current incremental data. In the next period, the model would have a performance degradation on samples containing stale features, which we call the feature staleness problem. To mitigate this problem, we propose a Feature Staleness Aware Incremental Learning method for CTR prediction (FeSAIL) which adaptively replays samples containing stale features. We first introduce a staleness aware sampling algorithm (SAS) to sample a fixed number of stale samples with high sampling efficiency. We then introduce a staleness aware regularization mechanism (SAR) for a fine-grained control of the feature embedding updating. We instantiate FeSAIL with a general deep learning-based CTR prediction model and the experimental results demonstrate FeSAIL outperforms various state-of-the-art methods on four benchmark datasets.