New Theory Cracks Open the Black Box of Deep Learning

By N. Wolchover. 2017

46 citations recorded by the source corpus.

Subfields: Health Informatics, Information Systems

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Health Informatics

1. Song-Chun Zhu
2. Yilan Chen
3. Xu Cheng
4. Yu Yang
5. Zhefan Rao
6. Quanshi Zhang
7. Ying Nian Wu

In Information Systems

1. Song-Chun Zhu
2. Yilan Chen
3. Xu Cheng
4. Yu Yang
5. Zhefan Rao
6. Quanshi Zhang
7. Ying Nian Wu

Citations in the corpus

• Unsupervised Learning of Neural Networks to Explain Neural Networks (2018, background): “Instead of analyzing network features from a global view [36, 26, 23], Bau et al.”
• Interpreting CNNs via Decision Trees (2018, background): “Although some theories, such as the information bottleneck [34], analyzed statistical characteristics of a neural”
• Interpretable Convolutional Neural Networks (2017, background): “spective of the information bottleneck [32] as follows.”
• Wireless Networks Design in the Era of Deep Learning: Model-Based, AI-Based, or Both? (2019, background): “A recent contribution in this direction is [204], which employs the so-called information bottleneck approach.”
• Explaining Knowledge Distillation by Quantifying the Knowledge (2020, methodology): “When learning from raw data, the DNN usually tries to model various visual concepts in early epochs and then discard non-discriminative ones in later epochs [41, 36], which leads to unstable optimization directions.”