lecture 1a introduction to ai n.
Skip this Video
Loading SlideShow in 5 Seconds..
Lecture 1A: Introduction to AI PowerPoint Presentation
Download Presentation
Lecture 1A: Introduction to AI

Lecture 1A: Introduction to AI

6 Views Download Presentation
Download Presentation

Lecture 1A: Introduction to AI

- - - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

  1. Lecture 1A: Introduction to AI Professor, Ruiyun Yu Software College, Northeastern University Email:

  2. Course overview

  3. TextBook • Deep Learning (Adaptive Computation and Machine Learning) , Ian Goodfellow. Yoshua Bengio & Aaron Courville. • Deep Learning with Python, Francois Chollet. • Artificial Intelligence: A Modern Approach (Third Edition), Stuart J. Russell & Perter Norvig.

  4. Outline • AI Introduction • What is AI? • Brief history of AI • Great persons • What’s the state of AI now? • Course overview • Mind mapping

  5. Your daily life AI Vision Could an intelligent agent living on your home computer manage your email, coordinate your work and social activities, help plan your vacations…… even watch your house while you take vacations.

  6. What is AI? Views of AI fall into four categories: The textbook advocates "acting rationally"

  7. Acting humanly: Turing Test • Turing (1950) "Computing machinery and intelligence": “Can machines think?”  “Can machines behave intelligently?” • Predicted that by 2000, a machine might have a 30% chance of fooling a lay person for 5 minutes • Anticipated all major arguments against AI in following 50 years • Suggested major components of AI: knowledge, reasoning, language understanding, learning • Capabilities • Natural language processing • Knowledge representation • Automated reasoning • Machine learning • Computer vision • robotics

  8. Thinking humanly: cognitive modeling • 1960s "cognitive revolution": information-processing psychology • Requires scientific theories of internal activities of the brain • How to validate? Requires • Predicting and testing behavior of human subjects (top-down) • Direct identification from neurological data (bottom-up) • Both approaches (roughly, Cognitive Science and Cognitive Neuroscience) are now distinct from AI

  9. Thinking rationally: "laws of thought" • Aristotle: what are correct arguments/thought processes? • Syllogisms: patterns for argument structures • Socrates is a man; all men are mortal; Socrates is mortal. • Logicians in the 19th century developed a precise notation for statements about all kinds of objects in the world Problems: • It’s not easy to take informal knowledge and state it in the formal terms required by logical notation • There is a big difference between solving a problem “in principle” and solving it “in practice”

  10. Acting rationally: rational agent • Rational behavior: doing the right thing • The right thing: which is expected to maximize goal achievement, given the available information The rational-agent approach has two advantages • It’s more general than the “laws of thought” because correct inference is just one of several possible mechanisms for achieving rationality • Second , it’s more amendable to scientific development than approaches based on human behavior or human thought. • One point to keep in mind: we will see before too long that achieving perfect rationality is not feasible in complicated environments

  11. Brief history of AI • In 1950, the famous Turing test was born, according to the definition of "the father of artificial intelligence" Alan Turing: if a machine can talk to humans (via telex equipment) without being identified as a machine, then the machine is said to be intelligent. • That same year, Turing also predicted the possibility of creating machines with real intelligence. Alan Mathison Turing

  12. Turing Test • Three rooms contain a person, a computer, and an interrogator. • The interrogator can communicate with the other two by teleprinter. • The interrogator tries to determine which is the person and which is the machine. • The machine tries to fool the interrogator into believing that it is the person. • If the machine succeeds, then we conclude that the machine can think.

  13. Brief history of AI • The first programmable robot was born in 1954. • In the summer of 1956, Dartmouth college in the United States held the first ever seminar on artificial intelligence, which was regarded as a sign of the birth of artificial intelligence.

  14. Brief history of AI • 1966 ~1972,the first artificial intelligence robot Shakey was born • The program actually provides some simulated responses to answer user questions. • It shows people that a program that is strong enough should be intelligent.

  15. Brief history of AI • In the early 1970s, artificial intelligence hit a bottleneck.The limited memory and processing speed of computers at the time were not enough to solve any real ai problems.(known as an "AI winter") • There's this stupid myth out there that AI has failed, but AI is around you every second of the day. • ——Rodney Brooks

  16. Brief history of AI • The term "winter of AI" was coined by researchers who experienced funding cuts in 1974.They note the craze for expert systems, and expect disappointment soon. Unfortunately for them, the usefulness of expert systems is limited to specific situations. • By the late 1980s, the new head of the defense advanced research projects agency (DARPA) thought ai was not "the next wave," and grants would favor projects that looked easier to produce.

  17. Brief history of AI • 1997: computer deep blue wins world chess championship. • 2011: developed an artificial intelligence program that answers questions using natural language • 2012:The Canadian team of neuroscientists created a virtual brain with simple cognitive abilities and 2.5 million simulated "neurons," named "Spaun," and passed the most basic IQ tests.

  18. Brief history of AI • 2013: deep learning algorithm is widely used in product development. • 2015: the year of breakthrough in artificial intelligence. • 2016: AlphaGo beats go world champion lee sedol. • 2017:AlphaGo Zero triumphed over AlphaGo.

  19. Great persons • Andrew Ng • Ng is one of the most authoritative international scholars in the field of artificial intelligence and machine learning. • Yann LeCun • He is well known for his work on optical character recognition and computer vision using convolutional neural networks (CNN), and is a founding father of convolutional nets.

  20. State of the Art • Machine learning • Computer vision • Natural language processing • Robotics • …

  21. Machine learning • How the computer simulates or implements human learning behavior in order to acquire new knowledge or skills, reorganize the existing knowledge structure and continuously improve its performance. • Deep Learning • Transfer Learning • Reinforcement Learning • …

  22. Computer vision • Computer vision is the science of how to make artificial systems "aware" from images or multidimensional data. • Image Classification • Object Localization • Semantic Segmentation • Instance Segmentation • …

  23. Natural language processing • It studies various theories and methods which can realize effective communication between human and computer by natural language. • Semantic analysis • Speech recognition • …

  24. Robotics • A robot is a machine that automatically executes the work. It can not only accept human command, but also run pre programmed programs, or act according to the principles and principles formulated by AI technology. • Automation • Reinforcement learning • …

  25. Course overview

  26. Search • Uniformed Search • Depth First Search • Breadth First Search • Uniform Cost Search • Informed Search • Greedy Search • A* Search • Adversarial Search • Games • Minimax • Alpha-beta pruning

  27. Probabilistic Inference • Probability Theory • Bayes Networks • Conditional Independence • Probabilities in BNs • A Common BN • Probabilistic Inference • Inference by Enumeration • Variable Elimination • Approximate Inference

  28. Decision Trees • Constructing a Decision Tree • ID3 • C4.5 • Regression Trees • CART • GBDT

  29. Clustering and Recommendation • Clustering • K-Means • EM • Mean-Shift • Spectral Clustering • Recommendation • Collaborative Filtering • Matrix Factorization

  30. Neural Networks • Perception • Depp Neural Network Structure • Back Propagation

  31. Convolutional Neural Network • Convolutional Neural Network • Convolution • Pooling • CNN Forward Propagation • CNN Backward Propagation • CNN Architectures • LeNet-5 • AlexNet • VGGNet • GoogleNet • ResNet

  32. Recurrent Neural Networks • Recurrent Neural Network • Vanilla RNNs • Some RNN Variants • Back Propagation through Time • Gradient Vanishing/Exploding • Long Short-term Memory • LSTM Neuron • Multiple-Layer LSTM • Back Propagation through Time in LSTM • Time-Series Prediction