This resource is no longer available
Data needs to be labeled and high-quality before you deploy any cutting-edge models or machine learning algorithms, but this can be difficult to do.
Active and semi-supervised learning can help you tackle unlabeled data and the problems it creates, but how much they help or hurt can be unclear.
Explore the results of this experiment to learn how semi-supervised learning and active learning function in various data labeling and data quality related use cases.