February 2024
Intermediate to advanced
200 pages
11h 56m
English
Loss function design, training algorithm, training data processing, and model architecture
In this chapter, we introduce the main techniques developed to mitigate the noisy-label learning problem. We present a hierarchical structure of the techniques, enabling us to provide a clearer explanation of the relationships between them. In addition to detailed explanations, we provide practical examples in simulated noisy-label learning problems. The understanding of this chapter is essential to facilitate the introduction of the state-of-the-art approaches that are presented in Chapter 5.
Noisy-label learning techniques; Noisy-label learning loss functions; Noisy-label learning algorithms; Noisy-label ...
Read now
Unlock full access