[機器學習入門] 李宏毅機器學習筆記-13 (Semi-supervised Learning ;半監(jiān)督學習)
VIDEO |
---|
Introduction
Why semi-supervised learning helps?
Semi-supervised Learning for Generative Model
Supervised Generative Model VS Semi-supervised Generative Model
Step
Why粉渠?
Low-density Separation
Self-training
Entropy-based Regularization
Outlook: Semi-supervised SVM
Smoothness Assumption
核心思想:近朱者赤污呼,近墨者黑
Classify astronomy vs. travel articles
更多的數(shù)據(jù)連在一起盾碗,很難分類,那么如何做呢桃焕?
Cluster(群集 ) and then Label
這種方法不一定made sense ,需要class很強乖菱。
But乖寒,How to know x1 and x2 are close in a high density region (connected by a high density path)
還有另一種方法:
Graph-based Approach
Graph Construction
怎樣在Graph 中定量地表示平滑度
將該式子整理一下,換個形式
如此院溺,讓smoothness 影響Loss楣嘁,as a regularization term
smoothness不一定要放在output上,放到任何一層都可以珍逸。
Better Representation
去蕪存菁逐虚,化繁為簡
Looking for Better Representation