How can we build systems that perform well in unknown environments and unforeseen situations? How can we develop systems that exhibit “intelligent” behavior, without prescribing explicit rules? How can we build systems that learn from experience in order to improve their performance? We will study core modeling techniques and algorithms from statistics, optimization, planning, and control and study applications in areas such as sensor networks, robotics, and the Internet. The course is designed for upper-level undergraduate and graduate students.
- Due date of final project postponed to December 7.
- We created a new mailing list (cs154). If you're taking the class but not on this list, please email the TAs.
- Andreas Krause (krausea [at] caltech.edu)
- Teaching Assistants:
- Pete Trautman (trautman [at] cds.caltech.edu)
- Xiaodi Hou (xiaodi.hou [at] gmail.com)
- Liang Wang (lwang2 [at] caltech.edu)
- Carlos R. Gonzalez (crgonzal [at] caltech.edu)
- Time and Location: Fall
& Wed 2:30pm-4pm in 105 Annenberg
- Prerequisites: Basic knowledge in probability and statistics (Ma 2b or equivalent) and algorithms (CS 1 or equivalent)
- 9 Units (3-0-6): Grading based on
- Homework assignments (50 %)
- Final exam (20 %)
- Challenge project (30 %)
- Office hours / recitation:
- Office hours TAs: TAs Tuesday, 4:30pm-5:30pm, Annenberg 107
- Office hours Instructor: Wednesday 4pm-5pm, Annenberg 300
- Recitation (optional but encouraged): Thursday 4:30pm-5:30pm, Annenberg 107
- Collaboration policy:
- Homeworks: Discussion of the problems is allowed, however, everybody must turn in their own solutions.
- Challenge problem: Groups of 2-3 students (exceptions possible with instructor's permission)
- Exam: Open textbook and lecture notes, but
no collaboration and external material allowed
- Stuart Russell, Peter Norvig. Artificial Intelligence: A Modern Approach (3rd edition). Prentice Hall (Dec 2009)
- Late day policy: Everybody has 3 late days that can be used for any homework assignment (e.g., you can be later for HW2 for 2 days and HW1 for 1 day). No credits when turning in after the late days have been used. Start early!
- Out: December 8, 10am
- Due: December 9, 10am
- Open textbook and lecture notes, but no collaboration and external material allowed
- Review in recitation on December 2
Challenge and independent projects
- Groups formed and project preferences selected: Oct 11
- For independent projects: Proposal due Oct 11
- Project milestone due Nov 3. You should have received a link to code per email. If not, please contact the TAs.
- Project final implementation due Dec 7
HomeworksHomeworks submitted by email should be sent to your TAs.
- Homework 1 out [pdf]: Oct 10, due: Oct 22 5pm in Lisa Knox's office, 343 Annenberg
- Homework 2 out [pdf]: Oct 29, due: Nov 10 5pm in Lisa Knox's office, 343 Annenberg
- Homework 3 out [pdf]: Nov 14, due: Nov 24 5pm in Lisa Knox's office, 343 Annenberg
- Sept 27 - Introduction. Chapters 1 and 2.
- Sept 29 - Uninformed search. Chapters 3.1-3.4.
- Oct 4 - Informed search. Chapters 3.5-3.6.
- Oct 6 - Adversarial search. Chapters 5.1-5.5.
- Oct 11 - Partial observability; CSPs. Chapters 4.3-4.4; 6.1
- Oct 13 - CSPs continued. Chapters 6.2-6.5
- Oct 18 - Propositional Logic. Chapters 7.1-7.5
- Oct 20 - Logical inference. First order logic.. Chapters 7.5, 8.1-8.3
- Oct 25 - FOL continued. Chapters 8.1-8.3
- Oct 27 - Probability. Chapters 13.1-13.3
- Nov 1 - Bayesian Networks. Chapters 14.1-14.2 JavaBayes applet
- Nov 3 - Bayesian Network inference: Variable elimination. Chapter 14.4
- Nov 8 - Approximate inference in Bayesian Networks. Chapter 14.5; Bishop Chapter 8.4
- Nov 10 - Information gathering. Chapters 16.3, 16.6, Bishop Chapter 1.6
- Nov 15 - Temporal models. Chapters 15.1-15.5
- Nov 17 - Markov Decision Processes. Chapters 17.1-17.3 Value Iteration demo (at UBC)
- Nov 22 - Learning: Regression and Classification. Chapter 18.1-2, 18.6; Bishop Chapter 1.1, 3.1. Linear regression demo (at UIUC). Logistic regression demo (at Technion)
- Nov 24 - Learning Bayesian Networks. Chapter 20.1-2
- Nov 29 - Reinforcement learning. Chapter 21.1-3 Q-learning demo (at Northwestern).
- Dec 1 - Probabilistic first order logic. Chapter 14.6 Alchemy (University of Washington)
- Christopher M. Bishop. Pattern Recognition and Machine Learning. Springer, 2007 (optional)