Criteo (Display Advertising Challenge)

Criteo包含7天的点击数据，广泛用于CTR预测基准测试。Criteo数据集中有26个匿名分类字段和13个连续字段。显示广告是十亿美元的努力，也是机器学习在互联网上的主要用途之一。但是，其数据和方法通常保持在锁和钥匙下。在这项研究竞赛中，CriteoLabs将分享一周的数据，供您开发预测广告点击率 (CTR) 的模型。给定一个用户和他正在访问的页面，他点击给定广告的概率是多少？这项挑战的目标是为CTR估计提供最准确的ML算法基准。

Criteo contains 7 days of click-through data and is widely used as a benchmark for click-through rate (CTR) prediction tasks. The Criteo dataset consists of 26 anonymized categorical features and 13 continuous features. Display advertising is a multi-billion-dollar industry and one of the primary applications of machine learning on the Internet. However, its data and related methodologies are usually kept confidential. In this research competition, CriteoLabs will share one week’s worth of data for participants to develop models that predict advertising CTR. Given a user and the webpage they are visiting, what is the probability that the user will click on a given advertisement? The goal of this challenge is to provide a benchmark for the most accurate machine learning algorithms for CTR estimation.

搜集汇总

背景与挑战

背景概述

Criteo数据集是一个用于点击率(CTR)预测基准测试的公开数据集，包含7天的点击数据，具有26个匿名分类字段和13个连续字段。该数据集由Criteo Research于2014年发布，旨在为CTR估计提供最准确的机器学习算法基准。