Course Information

Introduction to Intelligent Systems - Fall 2020

Instructor: Prof. Songhwai Oh (오성회)
Email: songhwai (at) snu.ac.kr
Office Hours: Friday 2:00-4:00PM
Office: Building 133 Room 405
Course Number: 430.457
Time: MW 11:00-12:15 PM
Location: Online (Building 301 Room 103)

TA-1: Jeongho Park (박정호)
Email: jeongho.park (at) rllab.snu.ac.kr
Office: Building 301 Room 814

TA-3: Wooseok Oh (오우석)
Email: wooseok.oh (at) rllab.snu.ac.kr
Office: Building 301 Room 814

TA-2: Hogun Kee (기호건)
Email: hogun.kee (at) rllab.snu.ac.kr
Office: Building 301 Room 814

TA-4: Jae Eun Kim (김재은)
Email: jaeeun.kim (at) rllab.snu.ac.kr
Office: Building 301 Room 718

Course Description

This course introduces the foundations of intelligent systems, such as probabilistic modeling and inference, statistical machine learning, computer vision, and robotics, to undergraduate students. Topics include Bayesian networks, hidden Markov models, Kalman filters, Markov decision processes, linear regression, linear classification, neural networks, deep learning, nonparametric models, and reinforcement learning. Students will also learn about how these methods are applied to practical applications such as computer vision and robotics. Lectures will be in English.

RC Car Racing: A Navigation Challenge

Project and Announcements

  • Project Information
  • [12/10] Hardware setup (12/14-12/17) 
  • [12/09] Final Project (due: 12/13, 23:59 KST)
  • [11/24] Project 4 (due: 12/04, 23:59 KST) 
  • [11/11] Project 3 (due: 11/23, 23:59 KST) 
  • [10/28] Project 2 (due: 11/11, 23:59 KST) 
  • [10/13] The midterm will be held in class on 10/21 (Wed). The exam is closed-book but you can bring one sheet (A4) of hand written notes on a single side (the other side must be blank). You have to turn in this cheat sheet with your exam. 
  • [09/28] Project 1 (due: 10/14, 23:59 KST) 
  • [09/16] Preliminary Project 2 (due: 9/28, 23:59 KST)
  • [09/02] Preliminary Project 1 (due: 9/16, 23:59 KST)
  • [08/28] Please read the Ethics of Learning.

Schedule

Week
Reading
Date
Lecture
Date
Lecture
1
 
 9/2
  •  Introduction
9/4
  • Makeup lecture
  • Time: 5-8 PM
  • Location: Online
2
AIMA Ch. 13, Ch. 14.1 - 14.3
9/7
  • Traditional AI
  • Review of probability
9/9
  • Bayesian networks
3
AIMA Ch. 14.4 - 14.5
9/14
  • Exact inference in Bayesian networks
9/16
  • Approximate inference in Bayesian networks
4
AIMA Ch. 15.1 - 15.3
9/21
  • Dynamic models
  • Inference in dynamic models
9/23
  • Hidden Markov models
5
AIMA Ch. 15.3 - 15.6
9/28
  • Kalman filtering
  • Kalman filtering (vector case)
9/30
  • Holiday
 
 
 
 
 
  • Makeup lecture
  • Dynamic Bayesian networks
6
AIMA Ch. 18.1 - 18.3, 18.4
10/5
  • Supervised learning
  • Decision trees
10/7
  • Decision trees
  • Generalization error
7
AIMA Ch. 18.6
10/12
  • Linear regression
10/14
  • Linear classification
8
AIMA Ch. 18.7
10/19
  • Artificial neural networks
10/21
  • Midterm
  • Time: 11:00-12:15 PM
  • Location: 301-302
9
 
10/26
  • Deep learning (intro)
10/28
  • CNN (part 1)
  • CNN (part 2)
  • RNN
10
AIMA Ch. 18.8 - 18.9
11/2
  • Nonparametric models
11/4
  • Support vector machines
11
AIMA Ch. 20.1 - 20.3
11/9
  • Bayesian learning
  • Learning with complete data
11/11
  • EM Algorithm
12
AIMA Ch. 16
11/16
  • Utility theory
11/18
  • Decision networks
13
AIMA Ch. 17.1 - 17.4, Ch. 21
11/23
  • Markov decision processes
  • POMDPs
11/25
  • Reinforcement learning
14
AIMA Ch. 25
11/30
  • Deep reinforcement learning
12/2
  • Robotics (intro)
15
AIMA Ch. 25
12/7
  • Localization
  • Mapping
12/9
  • SLAM
16
AIMA Ch. 24
12/14
  • Path planning
12/16
  • Computer vision: applications
  • Computer vision
 
 
 
 
12/19

Textbook

  • [Required] Stuart Russell and Peter Norvig. Artificial Intelligence: A Modern Approach (3rd edition), Prentice Hall, 2009. (AIMA Website)

Topics

  • Review of probability and linear algebra
  • Probabilistic Modeling and Inference:
    • Bayesian networks, Hidden Markov models, Kalman filters
    • Markov decision processes
  • Machine Learning:
    • Linear classification, Linear regression, Learning with complete data
    • Deep learning
    • Learning with hidden variables, EM algorithm
    • Nonparametric models, Support vector machines
    • Reinforcement learning
  • Robotics:
    • Localization and mapping, Motion planning, Planning uncertain movements, Moving
    • Robotic software architectures, Application domain

  • Computer Vision:

    • Image formation, Edge detection, Texture, Optical flow, Image segmentation
    • Object recognition, Reconstructing the 3D world