This outline is still very preliminary, and is subject to change.  We'll put the latest updates to the syllabus here when they become available.

Introduction & Logic [Geoff ~3 Lectures]

Reading: Russell & Norvig Chapters 7-9
  • Propositional Logic
  • First-order Logic
  • Proofs
    • resolution
    • unification
  • Satisfiability

Tues., Jan. 13:

Thurs., Jan. 15:

Tue., Jan. 20:

Thurs., Jan. 22:

[Top]

Search & Optimization [Tuomas ~9 Lectures]

Topics:
  • Theorem Proving as Search
  • Constraint Satisfaction Problems
    • variable and value ordering heuristics
    • forward checking
    • arc consistency
    • conflict-directed backjumping
    • constraint/clause learning
    • cutsets
    • tree decomposition
    • phase transitions
    • iterative refinement
    • DPLL
  • Optimization
    • informed search (i.e. with an objective and h-function)
    • MIP including Gomory mixed integer cut
    • pseudo-boolean inequalities
    • relating CSP, (Max)SAT, and MIP
    • binary search on objective value
    • relating clause learning & cuts
    • relating resolution, LP relaxation, and cuts
    • Optional Topic: RRTs/spatial search (Geoff)

Tues., Jan. 27:

Thurs., Jan. 29:

Tues., Feb. 3:

Thurs., Feb. 5:

Readings:

Tues., Feb. 10:

Thurs., Feb. 12:

Readings:

Tues., Feb. 17:

Thurs., Feb. 19:

[Top]

Linear Programs, Convex Programs, Duality [Geoff ~2 Lectures]

Reading:

Tues., Feb. 24:

Thurs., Feb. 26:

Optional Readings:

[Top]

Planning [Geoff ~2 Lectures]

Reading: Russell & Norvig Ch 11-12
  • Partial order planning
  • Graphplan / SATplan

Tues., March 3rd:

Thurs., March 5th:

[Top]

Probabilistic Inference [Geoff ~4 Lectures]

Reading: Topics:
  • Basic Probability
    • marginals
    • conditionals
    • integration/summation
  • Maximum Likelihood
    • relation to optimization
    • partition function
    • natural/expectation parameters
  • Undirected Graphical Models (MRFs)
    • relation to SAT
    • variable elimination
    • factor graph inference
  • Gradient Descent Learning
  • MCMC
    • MH
    • Gibbs
    • particle filters
  • Optional Topics (if we have time)
    • HMMs/POMDPS/RL
    • EM algorithm

Tues., March 9th:

Tues., March 17th:

Thurs., March 19th:

Tues., March 24th:

Thurs., March 26th:

Tues., March 31th:

[Top]

Game Playing, Equilibrium Finding [Tuomas ~6 Lectures]

  • Game Types, Normal Form, Extensive Form, Solution Concepts, Solution Complexity
  • Algorithms for Solving Normal Form Games
    • Lemke-Howson
    • PNS
    • MIP Nash
    • LP for zero sum games
  • Complete Information Sequential Games
    • Case: Deep Blue for chess
  • Incomplete Information Sequential Games
    • sequence form
    • Gameshrink for automated (lossless) abstraction
    • Case: Poker
  • Correlated Equilibrium (Geoff)

Tues., April 21st:

Thurs., April 23rd:

Tues., April 28th:

[Top]

Multiagent Learning [Geoff ~2 Lectures]

  • No-external-regret
  • No-Phi-regret

[Top]

Mechanism Design [Tuomas 1 Lecture]

  • Mechanism Design Setting
  • Gibbard-Satterthwaite Impossibility
  • Vickrey-Clarke-Groves Mechanism
  • Auctions
  • Automated Mechanism Design

[Top]