轉(zhuǎn)載自http://baojie.org/blog/2013/01/27/deep-learning-tutorials/
?
Stanford Deep Learning wiki: http://deeplearning.stanford.edu/wiki/index.php/Main_Page
?
幾個(gè)不錯(cuò)的深度學(xué)習(xí)教程，基本都有視頻和演講稿。附兩篇綜述文章和一副漫畫(huà)。還有一些以后補(bǔ)充。
Jeff Dean 2013 @ Stanford
http://i.stanford.edu/infoseminar/dean.pdf
一個(gè)對(duì)DL能干什么的入門(mén)級(jí)介紹，主要涉及Google在語(yǔ)音識(shí)別、圖像處理和自然語(yǔ)言處理三個(gè)方向上的一些應(yīng)用。參《Spanner and Deep Learning》（2013-01-19）
Hinton 2009
A tutorial on Deep Learning
Slides http://videolectures.net/site/normal_dl/tag=52790/jul09_hinton_deeplearn.pdf
Video http://videolectures.net/jul09_hinton_deeplearn/? (3 hours)
從神經(jīng)網(wǎng)絡(luò)的背景來(lái)分析DL，為什么要有DL說(shuō)得很清楚。對(duì)DL的基本模型結(jié)構(gòu)也說(shuō)得很清楚。十分推薦
更多Hinton的教程 http://www.cs.toronto.edu/~hinton/nntut.html
斯坦福的Deep Learning公開(kāi)課（2012）Samy Bengio, Tom Dean and Andrew Ng
http://openclassroom.stanford.edu/MainFolder/CoursePage.php?course=DeepLearning
教學(xué)語(yǔ)言是Matlab。
參2011年的課程CS294A/CS294W? Deep Learning and Unsupervised Feature Learning
更多的斯坦福工作： Deep Learning in Natural Language Processing
Deep Learning for Natural Language Processing, 2009 tutorial by Ronan Collobert (senna author)?
-
http://ronan.collobert.com/pub/matos/2009_tutorial_nips.pdf
-
video http://videolectures.net/nips09_collobert_weston_dlnl/
這個(gè)介紹了DL在三個(gè)方向上的應(yīng)用：tagging (parsing), semantic search, concept labeling
Ronan Collobert的Senna是一個(gè)c的深度學(xué)習(xí)實(shí)現(xiàn)，只有2000多行代碼
ACL 2012 tutorial
Deep Learning for NLP (without Magic)?
http://www.socher.org/index.php/DeepLearningTutorial/DeepLearningTutorial?
by Richard Socher, Yoshua Bengio and Chris Manning?
PDF: http://nlp.stanford.edu/~socherr/SocherBengioManning-DeepLearning-ACL2012-20120707-NoMargin.pdf?
Video: http://www.youtube.com/watch?v=IF5tGEgRCTQ&list=PL4617D0E28A5781B0
Kai Yu’s Tutorialhttp://vipl.ict.ac.cn/News/academic-report-tutorial-deep-learning-dr-kai-yu?
On November 26, 2012
Title: “A Tutorial on Deep Learning”?
Abstract:?
In the past 30 years, tremendous progress has been made in building effective classification models. Despite the success, we have to realize that, in major AI challenges, the key bottleneck is not the quality of classifiers but that of features. Since 2006, learning high-level features using deep architectures has become a big wave of new learning paradigms. In recent two years, performance breakthrough was reported in both image and speech recognition tasks, indicating deep learning are not something ignorable. In this talk, I will walk through the recent works and key building blocks, e.g., sparse coding, RBMs, auto-encoders, etc. and list the major research topics, including modeling and computational issues. In the end, I will discuss what might be interesting topics for future research.?
Bio of Dr. Kai Yu:?
余 凱任百度技術(shù)副總監(jiān)，多媒體部負(fù)責(zé)人，主要負(fù)責(zé)公司在語(yǔ)音，圖像，音頻等領(lǐng)域面向互聯(lián)網(wǎng)和移動(dòng)應(yīng)用的技術(shù)研發(fā)。加盟百度前，余凱博士在美國(guó)NEC研究院擔(dān) 任Media Analytics部門(mén)主管(Department Head)，領(lǐng)導(dǎo)團(tuán)隊(duì)在機(jī)器學(xué)習(xí)、圖像識(shí)別、多媒體檢索、視頻監(jiān)控，以及數(shù)據(jù)挖掘和人機(jī)交互等方面的產(chǎn)品技術(shù)研發(fā)。此前他曾在西門(mén)子公司任Senior Research Scientist。2011年曾在斯坦福大學(xué)計(jì)算機(jī)系客座主講課程“CS121: 人工智能概論”。他在NIPS, ICML, CVPR, ICCV, ECCV，SIGIR, SIGKDD，TPAMI，TKDE等會(huì)議和雜志上發(fā)表了70多篇論文，H-index=28，曾擔(dān)任機(jī)器學(xué)習(xí)國(guó)際會(huì)議ICML10, ICML11, NIPS11, NIPS12的Area Chair. 2012年他被評(píng)為中關(guān)村高端領(lǐng)軍人才和北京市海聚計(jì)劃高層次海外人才。?
Slides link: http://pan.baidu.com/share/link?shareid=136269&uk=2267174042[1]?
Video link: KaiYu_report.mp4 (519.2 MB)?
Theano Deep Learning Tutorial這個(gè)是實(shí)戰(zhàn), 如何用Python實(shí)現(xiàn)深度學(xué)習(xí)
http://deeplearning.net/tutorial/
Code https://github.com/lisa-lab/DeepLearningTutorials?
Survey Papers很多，不過(guò)初學(xué)看這兩篇應(yīng)該就夠了
Yoshua Bengio, Aaron Courville, Pascal Vincent. (2012) Representation Learning: A Review and New Perspectives
Yoshua Bengio?(2009). Learning Deep Architectures for AI.
更多
- Itamar Arel, Derek C. Rose, and Thomas P. Karnowski. (2010)?Deep Machine Learning – A New Frontier in Artificial Intelligence Research ?這篇沒(méi)什么公式，也不長(zhǎng)，就是籠統(tǒng)的介紹一下
- 截至2009的一些重要文章 http://www.iro.umontreal.ca/~lisa/twiki/bin/view.cgi/Public/ReadingOnDeepNetworks
Deep Learning雖好，也要牢記它的局限
