《机器学习简介课件ppt.ppt》由会员分享，可在线阅读，更多相关《机器学习简介课件ppt.ppt（44页珍藏版）》请在三一办公上搜索。
1、Machine Learning：An Overview,石立臣,1,Outline,What is machine learning(ML)Types of machine learningWork flowPopular modelsApplicationsFutures,2,What is machine learning,Training set(labels known),Test set(labels unknown),f()=“apple”f()=“tomato”f()=“cow”,3,What is machine learning,DefinitionMachine lear
2、ning refers to a system capable of the autonomous acquisition and integration of knowledgeMachine learning is programming computers to optimize a performance criterion using example data or past experience,Computer,Data,Algorithm,Program,Knowledge,Knowledge(new),4,What is machine learning,Every mach
3、ine learning algorithm has three componentsRepresentationModel(rules,statistics,instance;logic,KNN,SVM,DNN,)EvaluationPerformance(accuracy,mse,energy,entropy,)OptimizationParameters Combinatorial optimizationConvex optimizationConstrained optimization,5,Types of machine learning,Supervised learningT
4、raining data includes desired outputsUnsupervised learningTraining data does not include desired outputsSemi-supervised learningTraining data includes a few desired outputsReinforcement learningRewards from sequence of actions,6,Types of machine learning,Supervised learningClassification:discrete ou
5、tputRegression:continuous output,Bias-variance,7,Training and Validation Data,Full Data Set,Training Data,Validation Data,Idea:train eachmodel on the“training data”and then testeach modelsaccuracy onthe validation data,8,Underfitting&Overfitting,PredictiveError,Model Complexity,Error on Training Dat
6、a,Error on Test Data,Ideal Rangefor Model Complexity,Overfitting,Underfitting,9,Types of machine learning,Unsupervised learningClusteringDimensionality reductionFactor analysis,10,Types of machine learning,Semi-supervised learningClustering or classification,11,Types of machine learning,Reinforcemen
7、t learningRobot&control,12,Work flow,Prediction,Training Labels,Training,Training,Image Features,Image Features,Testing,Test Image,Learned model,Learned model,Slide credit:D.Hoiem and L.Lazebnik,13,Work flow,Features,14,Work flow,ModelsLogic,RulesStatistical,Black box modelStatic,dynamic modelOnline
8、 learningEnsemble learning,15,Work flow,Architecture,Model,Feature,Hardware,16,Popular models,Linear model:logistic regression,linear discriminant analysis,linear regression(with basis function),17,Popular models,Nearest neighborFeature&distance,18,Popular models,Support vector machine,19,Popular mo
9、dels,Artificial neural network,20,Popular models,Decision tree,21,Popular models,Collaborative filtering,22,Popular models,Hierarchical clusteringK-meansSpectral clusteringManifold learning,23,Popular models,Hidden markov modelConditional random fields,24,Applications,25,Applications,26,Applications
10、,27,Applications,28,Applications,29,Applications,30,Applications,31,Applications,32,Applications,Attention,33,Applications,Image classification,34,Applications,35,Applications,Brain machine interface,36,Applications,37,Applications,38,Applications,39,Applications,40,Applications,Indirect illuminationRegression,41,Applications,Indirect illuminationkd-tree,42,Applications,The core is the data set!Othersfeaturesmodel&optimization,43,Futures,DecisionControlKnowledgePrediction,44,
