博文
Coursera: Neural Networks for ML- Lecture 10
|||
Ways to improve generalization
ICML2012_Bayesian Posterior Sampling via Stocastic Gradient Fisher Scoring: 这篇讲的是full bayesian learning with mini-batches
JMLR2014_Dropout-A Simple Way to Prevent Neural Networks from Overfitting
ICML2008_Extracting and Composing Robust Features with Denoising Autoencoders
2. Mixtures of Experts
3. The idea of full Bayesian learning
4. Making full Bayesian learning practical
5. Dropout
https://blog.sciencenet.cn/blog-1583812-816127.html
上一篇：Coursera: Neural Networks for ML- Lecture 9
下一篇：Coursera: Neural Networks for ML- Lecture 11
扫一扫，分享此博文