斯坦福大學(xué)機(jī)器學(xué)習(xí)課程講義第一講-機(jī)器學(xué)習(xí)介紹

1、IntroductionWelcomeMachine LearningSPAMMachine Learning Grew out of work in AI New capability for computers Examples: Database mining Large datasets from growth of automation/web. E.g., Web click data, medical records, biology, engineering Applications cant program by hand.E.g., Autonomous helicopte

2、r, handwriting recognition, most of Natural Language Processing (NLP), Computer Vision. Machine Learning Grew out of work in AI New capability for computers Examples: Database mining Large datasets from growth of automation/web. E.g., Web click data, medical records, biology, engineering Application

3、s cant program by hand.E.g., Autonomous helicopter, handwriting recognition, most of Natural Language Processing (NLP), Computer Vision. Self-customizing programsE.g., Amazon, Netflix product recommendations Understanding human learning (brain, real AI).IntroductionWhat is machine learningMachine Le

4、arningArthur Samuel (1959). Machine Learning: Field of study that gives computers the ability to learn without being explicitly programmed. Tom Mitchell (1998) Well-posed Learning Problem: A computer program is said to learn from experience E with respect to some task T and some performance measure

5、P, if its performance on T, as measured by P, improves with experience E. Machine Learning definitionClassifying emails as spam or not spam. Watching you label emails as spam or not spam. The number (or fraction) of emails correctly classified as spam/not spam. None of the abovethis is not a machine

6、 learning problem.Suppose your email program watches which emails you do or do not mark as spam, and based on that learns how to better filter spam. What is the task T in this setting? “A computer program is said to learn from experience E with respect to some task T and some performance measure P,

7、if its performance on T, as measured by P, improves with experience E.”Machine learning algorithms:Supervised learningUnsupervised learningOthers: Reinforcement learning, recommender systems. Also talk about: Practical advice for applying learning algorithms. IntroductionSupervised LearningMachine L

8、earningHousing price prediction. Price ($) in 1000sSize in feet2 Regression: Predict continuous valued output (price)Supervised Learning“right answers” givenBreast cancer (malignant, benign)ClassificationDiscrete valued output (0 or 1)Malignant?1(Y)0(N)Tumor SizeTumor SizeTumor SizeAgeClump Thicknes

9、sUniformity of Cell SizeUniformity of Cell ShapeTreat both as classification problems. Treat problem 1 as a classification problem, problem 2 as a regression problem. Treat problem 1 as a regression problem, problem 2 as a classification problem. Treat both as regression problems. Youre running a co

10、mpany, and you want to develop learning algorithms to address each of two problems.Problem 1: You have a large inventory of identical items. You want to predict how many of these items will sell over the next 3 months.Problem 2: Youd like software to examine individual customer accounts, and for eac

11、h account decide if it has been hacked/compromised. Should you treat these as classification or as regression problems? IntroductionUnsupervised LearningMachine Learningx1x2Supervised LearningUnsupervised Learningx1x2Source: Su-In Lee, Dana Peer, Aimee Dudley, George Church, Daphne KollerGenesIndivi

12、dualsOrganize computing clustersSocial network analysisImage credit: NASA/JPL-Caltech/E. Churchwell (Univ. of Wisconsin, Madison) Astronomical data analysisMarket segmentationCocktail party problemMicrophone #1Microphone #2Speaker #1Speaker #2Audio clips courtesy of Te-Won Lee.Microphone #1:Micropho

13、ne #2: Microphone #1:Microphone #2: Output #1:Output #2: Output #1:Output #2: Cocktail party problem algorithmW,s,v = svd(repmat(sum(x.*x,1),size(x,1),1).*x)*x);Source: Sam Roweis, Yair Weiss & Eero SimoncelliOf the following examples, which would you address using an unsupervised learning algorithm? (Check all that apply.) Given a database of customer data, automatically discover market segments and group customers into different market segments. Given email labeled as spam/not spam, learn a spam filter.Given a set o