Technical White Paper: Unpacking Data Labeling

By dataiku
Unpacking Data Labeling Semi-Supervised and Active Learning at Work While they are unique and each come with their own set of benefits and challenges, the data labeling techniques of semi-supervised learning (SSL) and active learning can be used to perform predictions and improve model accuracy. Gain an overview of three popular active learning packages and how they compare Explore if using the confidence of a semi-supervised model to select samples to be pseudo-labeled outperforms a fixed size selection or not Discover findings from a reproduction exercise based on results from the 2019 paper "Diverse Mini-Batch Active Learning"

 

Topics

  1. Machine_Learning (0.41)
  2. Backend (0.31)
  3. NLP (0.12)

Similar Articles

Introduction to Active Learning

An extensive overview of Active Learning, with an explanation into how it works and can assist with data labeling, as well as its performance and potential limitations. ...