Andrew機器學習課程 章節1——初識機器學習
問題1:什麼是機器學習以及機器學習的通用演算法
如今一般使用的是Tom Mitchell對機器學習的定義: 原文為:
A computer program is said to learn from experience E with respecet to some task T and some performance measure P,if its performance ont T, as measured by P,improves with experience E.
即:計算機程式從經驗E中學習,解決某一任務T,進行某一效能度量p,通過p測定在T上的表現因經驗E而提高
Machine learning algorithms:
1:supervised learning (監督學習)unsupervised learning (非監督學習)為主要的學習演算法
2: 此外還有reinforcement learning (強化學習),recommender systems(推薦系統)
supervised learning:
gives a dataset in whitch the "right answer" are given .
1.regression question(迴歸問題) continuous valued 目的:預測一個連續值的輸出
2.classification question(分類問題)Discrete valued 目的:預測離散值的輸出
unsupervised learning:
gives a dataset but not told what to do with it
clustering algorithm(聚類演算法):告訴資料集,將其分為不同的聚類
always used in:
1.organize computing clusters 2.social network analysis 3.market segmentation 4.Astronomicla data analysis