Source: thenewstack.io
Meta-Semi Is an AI Algorithm That ‘Learns How to Learn Better’
As a subset of machine learning, https://thenewstack.io/demystifying-deep-learning-and-artificial-intelligence/ is the form of artificial intelligence that is inspired by how human brains work. Now, a team of researchers from https://www.tsinghua.edu.cn/en/ are proposing an algorithm that would help deep learning AI models exploit the labeled data that is available more efficiently, without compromising too much on accuracy.
Pseudo-labeling is a technique used in semi-supervised learning where the model is initially trained with whatever labeled data is available. The trained model then predicts labels for the unlabeled data, thus creating a set of pseudo-labeled data.
With this approach, the team’s Meta-Semi algorithm was able to consistently perform better than other state-of-the-art semi-supervised algorithms, notably even with less labeled data and larger number of classes.
Pseudo-labeling is a technique used in semi-supervised learning where the model is initially trained with whatever labeled data is available. The trained model then predicts labels for the unlabeled data, thus creating a set of pseudo-labeled data.
With this approach, the team’s Meta-Semi algorithm was able to consistently perform better than other state-of-the-art semi-supervised algorithms, notably even with less labeled data and larger number of classes.
Related Articles
Community Partners
DevOps Careers