Composite Quantization for Approximate Nearest Neighbor Search (ICML 2014)該文第三頁左側，倒數第五行提到alternative optimization；NeNMF: An Optimal Gradient Method for Nonnegative Matrix Factorization, 該文第二頁，公式2上面兩行，block coordinate descent，以公式2和3為例；Feature Fusion Using Locally Linear Embedding for Classification提到的參考文獻Some Notes on Alternating Optimization;Two-Dimensional Linear Discriminant Analysis的第四頁提到的Due to the difficulty of computing the optimal L and R simultaneously, we derive an iterative algorithm in the following.
我個人理解，這幾個概念都是等價的。
‘alternating optimization’ or ‘alternative optimization’?
Sue (UTS) comment: ‘Alternating’ means you use this optimization with another optimization, one after the other. ‘Alternative’ means you use this optimization instead of any other.
我的GSM-PAF最后用的‘alternating optimization’