
Sebastian Raschka: AI Education and Research
The Gradient: Perspectives on AI
00:00
Ordinal Regression and its Application to Neural Networks
This chapter explores the concept of ordinal regression, which is a classification problem with labels that have a natural order. The speaker discusses the challenges of quantifying arbitrary distances between different categories and the subjectivity of assigning exact numbers to these differences. They also explain how to transform a pre-trained deep neural network into an ordinal regression model using the binary extension framework.
Transcript
Play full episode