Sign In

Communications of the ACM

ACM Opinion

Moving Towards Reproducible Machine Learning

View as: Print Mobile App Share:
Illustration shows a hand holding a light bulb inside of a network of ideas

Getty Images

The field of machine learning has faced a "reproducibility crisis" due to a lack of transparency and reporting surrounding the steps taken to build data-driven models.

As machine learning becomes more popular and widely used in many different fields, it becomes crucial to make sure that researchers are reporting all the details—about the collection, selection, and availability of datasets, models, and code—to allow proper reproducibility of the results.

From Nature Computational Science
View Full Article


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account