对于input的m个vectors xi∈Rnx_i\in\mathbb{R}^n, i∈[0,m−1]i\in[0, m-1] sparse coding : 找over-complete 的basis vector,k>nk>n, 升高维度, 并使得转换后的vector x^i\hat{x}_i尽量sparse,pca: 找 complete的basis vector, k<nk<n,
link: http://ufldl.stanford.edu/tutorial/ 1. linear rssification use MLE to understand the loss function
Copyright © 2005-2024 51CTO.COM 版权所有 京ICP证060544号