Clean Labels for ML
summary
Examines how the quality of ML models depends on the labels used for training and practical methods to improve labels collected from crowdsourcing platforms such as Toloka and Amazon Mechanical Turk. Describes concrete techniques and organizational approaches for collecting better labels, illustrated with real-world good and bad examples.
AI-written summary · about
chapters (5)
related talks
31mReliable Data for Large ML Models: Principles and Practices
20mGreenfield vs. Brownfield Data Labeling
04mManaging Machines
40mData labelling to scale your ML and AI pipelines
30mActive Learning and Human-in-the-Loop Machine Learning
29mLabeling Tools are Great, but What About Quality Checks?
02mDBTB INT Omar Alonso
30mMachine Learning: Beyond the Tutorial