Gabi Dobocanblog.telepat.io·Nov 24, 2024Unveiling DISCERN: A New Frontier for Bias Detection in Text ClassifiersExploring DISCERN: Overview and Main Claims DISCERN is a breakthrough framework that identifies and remedies systematic biases in text classifiers. It accomplishes this by generating natural language descriptions of errors, translating complex patter...Bias detection
DataChefforDataChef's Blogblog.datachef.co·Feb 7, 2022Assessing Pre-training Bias with AWS SageMaker ClarifyIntroduction In the previous article, we discussed why explaining machine learning (ML) models is as important as their performance and how it benefits all the stakeholders. Now, we are going to see what explainability and fairness of ML models requi...396 readsBias detection