018 Interpretability of Neural Networks (Lecture 7)

目录

• Outline
• I.A. Interpreting and visualizing Neural Networks with sliency maps
• I.B. Interpreting and visualizing Neural Networks with occusion sensitivity
• III.C. Interpreting NNs using class activation maps
• II.A. Visualizing NNs from the inside using gradient ascent(class model visualization)
• II. B. Visualizing NNs from the inside dataset search
• II.C. The decovolution and its applications
• III.Deep Dream
• Summary

Outline

I.A. Interpreting and visualizing Neural Networks with sliency maps

I.B. Interpreting and visualizing Neural Networks with occusion sensitivity

III.C. Interpreting NNs using class activation maps

II.A. Visualizing NNs from the inside using gradient ascent(class model visualization)

II. B. Visualizing NNs from the inside dataset search

II.C. The decovolution and its applications

III.Deep Dream

Summary