With enough data by John Norwood

With enough data

A narrow majority

Becomes certainty

 

This sciku has to do with probability theory used in data science and machine learning. In a nutshell, the more data you have, the lower the uncertainty of your model and the smaller the bias needed to reliably predict an eventual outcome. This paper by Dolev et al (2010) geeks out on some of the technical details of purifying data and machine learning.

Original research: https://doi.org/10.1145/1953563.1953567

John Norwood is a Mechanical Engineer working with Carbon, Inc. to revolutionize how things are made. His interests include old houses, yoga, baking, cryptography, and bluegrass music. You can follow him on Twitter under the handle @pryoga

Enjoyed this sciku? Check out some of John’s other work: The answer is none, God may be defined, Rivers cut corners, and Squeamish ossifrage.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.