11CS 6375 Machine LearningLecture 1: IntroductionInstructor: Yang LiuCS6375 Machine Learning2General information Where: ECSN 2.311 When: Mon and Wed 2:30-3:45pm Instructor: Yang Liu, ECSS 3.402 Email: [email protected] Office hour: M: 3:45-4:30pm; W: 1:30-2:30pm2General information Course website: http://www.hlt.utdallas.edu/~yangl/cs6375 TA: TBD Class notes and handouts: available from course web site Discussion and announcements: eLearningCS6375 Machine Learning3CS6375 Machine Learning4Textbooks Reference textbooks: Machine Learning, Tom Mitchell Pattern Recognition and Machine Learning, Christopher Bishop Introduction to Machine Learning, Ethem Alpaydin Artificial Intelligence, Russell and Novig Machine Learning: A Probabilistic Perspective, Kevin Murphy Other material available from the course webpage3CS6375 Machine Learning5Prerequisite CS 5345: Algorithm analysis and data structures Ability to program in C/C++, Java, or other languages Knowledge of math and probability/stats theory Ready to learnCS6375 Machine Learning6Tentative grading policy Homework assignments: 35% (some programming, some exercises) Midterm (March 4, tentative): 30% Final exam (UTD schedule): 30% Closed book exam, one cheat sheet allowed Quiz and class participation: 5%4CS6375 Machine Learning7Course policies Homework policy Collaboration is encouraged But you have to write your own solutions/programs Late assignment policy One day late: 85% 2 days late: 70% No assignment accepted after 2 days Re-grade policy Requests must be made within one week of when the work was returnedWhat is “Machine Learning” “Field of study that gives computers the ability to learn without being explicitly programmed” [ArtherSamuel] “A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E” [Tom Mitchell]CS6375 Machine Learning85CS6375 Machine Learning9What is “Machine Learning” Learn general models from a set of particular examples Data is cheap and abundant (sometimes may not be the case); knowledge is expensive and scarce E.g., use customer transactions to learn consumer behavior Build a model that is a good and useful approximation of the data The ability to perform a task in a situation which has never been encountered before (learning = generalization) Different learning: classification, planning, finding patterns, etc.CS6375 Machine Learning10Machine learning applications Speech recognition, natural language processing Computer vision Computational biology Association rules: somebody who buys X also buys Y Medical diagnosis: from symptoms to illness Security (detecting intrusion, worms, anomaly) More6CS6375 Machine Learning11Speech recognitionRecognize speechWreck a nice beachCS6375 Machine Learning12Text classification Spam filteringDear Sir/MadamI know that this letter may come to you as a surprise but due to the urgency of this transaction. First I must solicit your confidence in this transaction, this is by virtue of it's nature as being utterly confidential and top secret…..I am the manager of bill and exchange at the foreign remittance department of African Development bank (ADB). I came to know you in my private search for a reliable and reputable person to handle this Confidential Transaction, which involves the transfer of a huge sum of money to a foreign account requiring maximum confidence.I am writing to you, following the impressive information received about you from the chambers of commerce. I believed that you are capable and reliable to champion this business opportunity. In my department we discovered an abandoned sum of $30m US dollars (Thirty million US dollars). …..7Text classification Sentiment analysisTell me a movie that is more famous than this. Tell me a movie that has had more parodies spinned off its storyline than this. Tell me one movie that has been as quoted as a much as this. The answer is you can't. No movie has had as much of an impact as The Godfather has had ever since it was released.The acting was simply amazing, what else could you sayCS6375 Machine Learning13CS6375 Machine Learning14Ambiguity resolution in language Word selection (speech recognition) Can I have a peace of cake? Piece Word sense disambiguation …Nissan car and truck plant is .. Divide life into plant and animal kingdom Pronoun resolution The dog bit the kid. He was taken to a vet.8CS6375 Machine Learning15IBM Watson JeopardyIBM Watson Jeopardy “Using machine learning, statistical analysis and natural language processing to find and understand the clues in the questions, Watson compared possible answers, by ranking its confidence in their accuracy, and responded – all in about three seconds.”CS6375 Machine Learning169CS6375 Machine Learning17Fingerprint recognitiona) b) c)Image recognitionCS6375 Machine Learning18Note: using deep learning10CS6375 Machine Learning19Credit scoring Low-risk and high-risk customers based on income and savingsCS6375 Machine Learning20Play chess11CS6375 Machine Learning21Unmanned carCS6375 Machine Learning22Why study learning Important IT skills that employers look for (6-figure salary ☺) Lots of applications Computer systems with new capabilities Develop systems that are too difficult or impossible to construct manually Develop systems that can automatically adapt and customize themselves to the needs of the individual users through experience Discover knowledge and patterns in databases, data mining12CS6375 Machine Learning23Why study learning Understand human and biological learning Time is right Initial algorithms and theory in place Growing amounts of data Computational power available Budding industryCS6375 Machine Learning24Work in Machine Learning Artificial Intelligence Makes use of Probability and statistics, linear algebra, calculus, optimization Related to Philosophy, psychology, neurobiology, linguistics Has applications in AI (natural language, vision, planning, HCI) Computer science (compilers, systems, databases, software engineering, security)13CS6375 Machine Learning25Course overview Introduction to machine learning Supervised learning models
View Full Document
Unlocking...