Industry Insights from Comparing Deep Learning and GBDT Models for E-Commerce Learning-to-Rank

TL;DR

This study compares deep neural networks and traditional tree-based models for e-commerce ranking, showing DNNs can outperform tree models in clicks and revenue through systematic benchmarking and real-world testing.

Contribution

It provides the first comprehensive comparison of DNNs and LambdaMART in e-commerce ranking, demonstrating DNNs' superior performance in key metrics.

Findings

DNNs outperform LambdaMART in total clicks and revenue.

DNNs achieve parity in total units sold.

Results validated through 8-week online A/B testing.

Abstract

In e-commerce recommender and search systems, tree-based models, such as LambdaMART, have set a strong baseline for Learning-to-Rank (LTR) tasks. Despite their effectiveness and widespread adoption in industry, the debate continues whether deep neural networks (DNNs) can outperform traditional tree-based models in this domain. To contribute to this discussion, we systematically benchmark DNNs against our production-grade LambdaMART model. We evaluate multiple DNN architectures and loss functions on a proprietary dataset from OTTO and validate our findings through an 8-week online A/B test. The results show that a simple DNN architecture outperforms a strong tree-based baseline in terms of total clicks and revenue, while achieving parity in total units sold.