How to debug machine learning systems, according to Peter Norvig.
Originally shared by Wayne Radinsky
How to debug machine learning systems, according to Peter Norvig. "Instead of traditional test suite assertions that respond with true, false or equal, machine learning test assertions should respond with assessments, such as the results of today’s experiment were 90 percent good and consistent with tests run yesterday."
"Compounding the verification problem, the truth used to verify the product of a machine-learning system may not be known or distorted by human perception."
http://www.networkworld.com/article/3075413/software/google-ai-expert-explains-the-challenge-of-debugging-machine-learning-systems.html
How to debug machine learning systems, according to Peter Norvig. "Instead of traditional test suite assertions that respond with true, false or equal, machine learning test assertions should respond with assessments, such as the results of today’s experiment were 90 percent good and consistent with tests run yesterday."
"Compounding the verification problem, the truth used to verify the product of a machine-learning system may not be known or distorted by human perception."
http://www.networkworld.com/article/3075413/software/google-ai-expert-explains-the-challenge-of-debugging-machine-learning-systems.html
Comments
Post a Comment