Pengju

Transcription

1 Introduction to AI Pengju Ren Institute of Artificial Intelligence and Robotics 1

2 2

3 Self-Introduction Pengju Ren, Associate Professor Research topic: Novel Computer architecture for Machine Learning and Neuromorphic Computing. Personal website: Office: Room 211, Yifu Science Building 3

4 Teaching Assistant(TAs) 刘天良博士二年级 Focusing on Algorithm 马永强 博士三年级 Focus on Neuromorphic Computing 4

5 Teaching Assistant(TAs) 魏亚东研究生二年级 Focusing on Algorithm 赵之旭研究生二年级 Focus on GPU-based AI Applications (Self-driving Car/Drone) 樊珑研究生二年级 Focus on FPGA-based High Efficient AI computing Architecture 5

6 Recommended Textbook Artificial Intelligence-A Modern Approach(3 rd ) 6 The leading textbook in AI. Used in over 1300 universities in over 110 countries.

7 Other Textbooks Ian Goodfellow, Yoshua Bengio and Aaron Courville, Deep Learning,MIT Press book,2017 周志华著, 机器学习 清华大学出版社,2016 7

8 考核方式 闭卷考试成绩 50% 平时成绩 20% 实验作业 30% 8

9 Outline What is AI? AI Applications Modeling based AI AI Machine Learning and Deep Learning 9

10 What s AI? 10

11 What is AI? 11

12 What is AI? 12

13 What is AI? Artificial intelligence (AI) is intelligence exhibited by machines. In computer science, the field of AI research defines itself as the study of "intelligent agents": any device that perceives its environment and takes actions that maximize its chance of success at some goal. --- from Wikipedia Artificial intelligence is the search for a way to map intelligence into mechanical hardware and enable a structure into that system to formalize thought. No formal definition, as yet, is available for as to what artificial intelligence actually is. 13

14 What is AI? 14

15 Outline What is AI? AI Applications Modeling based AI AI Machine Learning and Deep Learning 15

16 AI is an enabling technology FINANCE ROBOTICS GAMES MEDICINE THE WEB 16

17 AI in Finance Use of AI in banking can be tracked back to 1987 when Security Pacific National Bank in USA set-up a Fraud Prevention Task force to counter the unauthorised use of debit cards. Banks use artificial intelligence systems to organize operations, maintain book-keeping, invest in stocks, and manage properties. AI can react to changes overnight or when business is not taking place. In August 2001, robots beat humans in a simulated financial trading competition. AI has also reduced fraud and crime by monitoring behavioral patterns of 17 users for any changes or anomalies. AI on Wall Street

18 AI in Robotics Intelligence is required for robots to be able to handle such tasks as object manipulation and navigation, with sub-problems of localization (knowing where you are, or finding out where other things are), mapping (learning what is around you, building a map of the environment), and motion planning (figuring out how to get there) or path 18 planning (going from one point in space to another point, which may involve compliant motion where the robot moves while maintaining physical contact with an object)

19 AI in Robotics 19

20 AI in GAMEs AlphaGo(AI) v.s. LEE SEDOL F.E.A.R. (Vivendi Universal*) use of scripted events is a type of AI Playing against Human player in adversarial game. Creating believable Characters in the Game. 20

21 AI in Medicine IBM Watson Precision medicine with AI Tech. Medication adherence Healthy behavior Support care givers Drug development 21

22 AI in Web server Understanding words you typed in the searching box. Finding the most relevant pages in a short period of time. 22

23 AI in Machine Translation 23

24 Autonomous Driving 24

25 Outline What is AI? AI Applications Modeling based AI AI Machine Learning and Deep Learning 25

26 Challenges of Modeling based AI? 26

27 Two complexities of AI (1) Computational Complexity : exponential explosion 27

28 Two complexities of AI (2) Information complexity: need to acquire knowledge 28

29 Challenges of Modeling based AI 29

30 Challenges of Modeling based AI 30

31 Challenges of Modeling based AI 31

32 Challenges of Modeling based AI 32

33 Challenges of Modeling based AI 33

34 Challenges of Modeling based AI Qualification Problem: is concerned with the impossibility of listing all the preconditions required for a real-world action to have its intended effect. Ramification Problem: refers to determining the indirect effects of actions. It might also be posed as how to represent what happens implicitly due to an action or how to control the secondary and tertiary effects of an action 34

35 Challenges of Modeling based AI 35

36 Outline What is AI? AI Applications Modeling based AI AI Machine Learning and Deep Learning 36

37 Machine Learning Approach The program produced by the learning algorithm may look very different from a typical hand-written program. It may contain millions of numbers. (about how you weight different kinds of evidence.) If we do it right, the program works for new cases as well as the ones we trained it on. If the data changes the program can change too by training on the new data (we should be able to change the program very easily, by retraining it on the new data) Massive amounts of computation are now cheaper than paying somebody to write a task-specific program. 37

38 AI > ML > DL Artificial Intelligence Human Intelligence Exhibited by Machines Machine Learning An Approach to Achieve Artificial Intelligence Deep Learning A Technique for Implementing Machine Learning 38

39 Machine Learning & AI 39

40 Machine Learning - Paradim 40

41 Machine Learning Example1 ImageNet Benchmarks,12.0 million testing pictures (with labels) 150,000 testing pictures, total of 1000 categories. 41

42 Machine Learning Example1 42

43 Machine Learning Example 2 43

44 Machine Learning Example 3 44

45 Machine Learning five tribes 45

46 Machine Learning - History 46

47 Machine Learning - Future 47

48 Machine Learning Methods(1) 48

49 Machine Learning Methods(2) 49

50 Machine Learning Methods(3) 50

51 Machine Learning Methods(4) 51

52 Machine Learning Methods(5) 52

53 Machine Learning Methods(6) 53

54 Machine Learning Methods(7) 54

55 Machine Learning Methods(8) 55

56 IoT Big Data Cloud and AI 56

57 Beyond Turing Test 57

58 Beyond Turing Test 58

59 Discuss Will Machine Replace Human being? 59