AI/Machine Learning Book List

現時AI好hit，未算係做到intelligence，而所謂既AI其實只係Machine learning (ML)
ML既foundations係由Applied Math, Statistics同CS組成
現時Deep learning (DL) 亦係非常hit，但係theoretical understanding極為不足，極似係black-box algorithm

所以如果想正式入手AI，應該係由ML fundamentals學起，而UC Berkeley既ML大師Prof. Michael Jordan曾經建議過一個book list俾ML既PhD學生


http://nzhiltsov.blogspot.hk/2014/09/highly-recommended-books-for-machine-learning-researchers.html
可以見到入面大部分既書都係statistics + applied math

希望係度拋下磚，等多D對AI/ML有興趣，而且math/stat又勁既連登仔去做ML/AI research
宜家入到top tech firm (Google/fb/Amazon/M$, 甚至大陸BAT) 做AI researcher應該有100K+/月

歡迎討論

留名，好想學ML 但唔太識上手

啲數學好撚深

你大學讀咩科先 如果全自學會好辛苦

讀CS 但係垃圾屎

啲數學我ok，弊在統計算法編程都好廢

其實好多library

Algorithm你基本明個concept就得

真係要寫嘅都係data processing

我唔係淨係想人明個concept就得
如果要淨係識用，其實唔係咩難事

咁咪再學囉，冇人天生識

I mean 唔使因為programming/cs廢而唔去學ml

因為真係寫code嘅部份係微不足道

要cs domain knowledge 嘅part寫library嗰班友已經幫你做哂

最近睇緊google個新ML course
都想睇翻d foundation 嘅materials

即係純ML theory嘅書？

Introduction to algorithm By T. H. Cormen, C. E. Leiserson, R. L. Rivest and C. Stein
呢本係classical 不過成舊磚咁厚

lm

Lm
D 數真係難
要明曬d數點黎要好多時間

Algorithms定ML Algorithms先

general algorithm
例如 headsort, quick sort

講梗上面巴打係問邊樣

lm

咁你quot埋我做咩

btw 巴打有ml algorithm書可推介？

Pattern Recognition and Machine Learning

留名

佢啲數真係又難又煩
試過為左想學深入啲原理走去take Math dept.既 Linear algebra 同multivariable calculus
結果爛grade收場
其實係咪反而practical啲學sklearn tensorflow 仲實際
求指點
利申CSmajor

如果你興趣只係一個比較窄D既AI field (例如deep neural network)，其實唔洗識哂果D math，講真其實都無人會識哂成個AI field，呢個field去得太快，你讀得哂果堆math書得黎已經又好多新野

落手做一次實際啲