Remove IT Remove KDD Remove Testing
article thumbnail

Experiment design and modeling for long-term studies in ads

The Unofficial Google Data Science Blog

A/B testing is used widely in information technology companies to guide product development and improvements. For questions as disparate as website design and UI, prediction algorithms, or user flows within apps, live traffic tests help developers understand what works well for users and the business, and what doesn’t.

article thumbnail

Density-Based Clustering

Domino Data Lab

Due to its importance in both theory and applications, this algorithm is one of three algorithms awarded the Test of Time Award at the KDD conference in 2014. The anomalous points pull the cluster centroid towards them, making it harder to classify them as anomalous points. neighborhoods. The general idea behind ?-neighborhoods

Metrics 116
article thumbnail

Using Empirical Bayes to approximate posteriors for large "black box" estimators

The Unofficial Google Data Science Blog

In the examples above, we might use our estimates to choose ads, decide whether to show a user images, or figure out which videos to recommend. These decisions are often business-critical, so it is essential for data scientists to understand and improve the regressions that inform them. The size and importance of these systems makes this hard.

KDD 40