斯坦福大學-機器學習課程Stanford Engineering Everywhere-MachineLearning 英文版 DVD 只能於電腦播放

相對於其他名校，斯坦福大學的工科課程更注重實用性。這也是我個人很讚賞的一點。
關於發布本資源的初衷。坦白的說，人工智能的發展到已經進入了一個瓶頸期。近年來各個研究方向都沒有太大的突破。真正意義上人工智能的實現目前還沒有任何曙光。但是，機器學習無疑是最有希望實現這個目標的方向之一。斯坦福大學的“StanfordEngineeringEverywhere”免費提供學校裡最受歡迎的工科課程，給全世界的學生和教育工作者。得益於這個項目，我們有機會和全世界站在同一個數量級的知識起跑線上。

此課程獻給所有同好。讓我們向著朝陽奔跑吧~

本課程來源於斯坦福大學的“StanfordEngineeringEverywhere”項目。
首頁為：http://see.stanford.edu/default.aspx
目前已有的課程是：
IntroductiontoComputerScience:
ProgrammingMethodologyCS106A
ProgrammingAbstractionsCS106B
ProgrammingParadigmsCS107


ArtificialIntelligence:
IntroductiontoRoboticsCS223A
NaturalLanguageProcessingCS224N
MachineLearningCS229


LinearSystemsandOptimization:
TheFourierTransformanditsApplicationsEE261
IntroductiontoLinearDynamicalSystemsEE263
ConvexOptimizationIEE364A
ConvexOptimizationIIEE364B

本課程為ArtificialIntelligence裡的MachineLearningCS229

課程簡介：
ArtificialIntelligence|MachineLearning
Instructor:Ng,Andrew

Thiscourseprovidesabroadintroductiontomachinelearningandstatisticalpatternrecognition.

Topicsinclude:supervisedlearning(generative/discriminativelearning,parametric/non-parametriclearning,neuralnetworks,supportvectormachines);unsupervisedlearning(clustering,dimensionalityreduction,kernelmethods);learningtheory(bias/variancetradeoffs;VCtheory;largemargins);reinforcementlearningandadaptivecontrol.
Thecoursewillalsodiscussrecentapplicationsofmachinelearning,suchastoroboticcontrol,datamining,autonomousnavigation,bioinformatics,speechrecognition,andtextandwebdataprocessing.
Studentsareexpectedtohavethefollowingbackground:

Prerequisites:-Knowledgeofbasiccomputerscienceprinciplesandskills,atalevelsufficienttowriteareasonablynon-trivialcomputerprogram.
-Familiaritywiththebasicprobabilitytheory.(Stat116issufficientbutnotnecessary.)
-Familiaritywiththebasiclinearalgebra(anyoneofMath51,Math103,Math113,orCS205wouldbemuchmorethannecessary.)

講師簡介：
AndrewNg

Ng'sresearchisintheareasofmachinelearningandartificialintelligence.HeleadstheSTAIR(STanfordArtificialIntelligenceRobot)project,whosegoalistodevelopahomeassistantrobotthatcanperformtaskssuchastidyuparoom,load/unloadadishwasher,fetchanddeliveritems,andpreparemealsusingakitchen.Sinceitsbirthin1956,theAIdreamhasbeentobuildsystemsthatexhibit"broadspectrum"intelligence.However,AIhassincesplinteredintomanydifferentsubfields,suchasmachinelearning,vision,navigation,reasoning,planning,andnaturallanguageprocessing.Torealizeitsvisionofahomeassistantrobot,STAIRwillunifyintoasingleplatformtoolsdrawnfromalloftheseAIsubfields.Thisisindistinctcontrasttothe30-year-oldtrendofworkingonfragmentedAIsub-fields,sothatSTAIRisalsoauniquevehiclefordrivingforwardresearchtowardstrue,integratedAI.
Ngalsoworksonmachinelearningalgorithmsforroboticcontrol,inwhichratherthanrelyingonmonthsofhumanhand-engineeringtodesignacontroller,arobotinsteadlearnsautomaticallyhowbesttocontrolitself.Usingthisapproach,Ng'sgrouphasdevelopedbyfarthemostadvancedautonomoushelicoptercontroller,thatiscapableofflyingspectacularaerobaticmaneuversthatevenexperiencedhumanpilotsoftenfindextremelydifficulttoexecute.Aspartofthiswork,Ng'sgroupalsodevelopedalgorithmsthatcantakeasingleimage,andturnthepictureintoa3-Dmodelthatonecanfly-throughandseefromdifferentangles.

資源中materials.rar是講義，當然也是英文的。