What is computation? What can be computed in principle with unbounded computational resources? What can be computed efficiently? What can we gain by formally modeling computation and how do different models relate to one another? How can models highlight different resources of computations, some obvious (such as time and memory) and others less so (such as communication and randomness). What is gained by considering natural and social phenomenon as computations and looking at central notions such as proofs, knowledge, learning, games, randomness, entropy and more through the computational lens?
We will consider these questions and others using a rigorous mathematical approach. We will discuss what we know as well as some of the central open problems in pure and applied mathematics, and specifically the P vs. NP problem.
Some specific topics: Finite Automata – Very Simple Models (constant memory), Non-determinism (power of guessing), Learning, communication complexity, Streaming algorithms, Powerful models – Turing Machines, Decidability, Kolmogorov Complexity, Time complexity, P vs. NP, NP-completeness, Other Resources: space, randomness, communication, power, … Crypto, Game Theory, … The Computational Lens.
Prerequisites: CS 103 or 103B.
Current Offering: Fall 2022 – Under Construction
Instructor: Omer Reingold, Gates 462, reingold (at stanford dot edu)
CAs:
Callum Jan Burgess, callumb (at stanford dot edu)
Christie Di, zidi (at stanford dot edu)
Felipe Correa Dias Godoy, fgodoy (at stanford dot edu)
Saumya Goyal, saumg (at stanford dot edu)
Arthur Lee, arthurlh (at stanford dot edu)
Location and Times:
Tue and Thursday, 10:30 AM – 11:50 PM at TBD.
The class is flipped. Pre-recorded lectures on YouTube (unfortunately contains ads) and Canvas. Also see Grid of Welcome, curtesy of Logan Matthew Bell and Ricardo Iglesias. A related and multi-instotutes project is here (could be a useful resource).
The Tuesday meetings will be devoted to Q&A as well as additional examples and will be posted on Canvas only. The Thursday meetings will mostly serve as OHs, not recorded (and not mandatory). In addition, Thursdays will be used for midterms and make-up classes.
Book: Michael Sipser, introduction to the theory of computation (2nd or 3rd edition)
– Extra reading: Boaz Barak, Introduction to Theoretical Computer Science (the approach is different from Sipser, but some parts could augment your understanding). Additional reading below.
Office Hours: See on Piazza
HW assignments (see on Piazza):
Homework will be assigned almost every Tuesday and will be due one week later at the beginning of class. No late submission. We will drop your lowest homework grade.
You may (even encouraged to) collaborate with others, but you must:
1. Try to solve all the problems by yourself first
2. List your collaborators on each problem
3. If you receive a significant idea from somewhere, you must acknowledge that source in your solution.
4. Write your own solutions (important!)
Assignments and submissions through gradescope.com
Best to write in LaTex
Lectures (future schedule is tentative)
Tuesday 9/27: Introduction
Assignment: Watch videos 1,2,4.
What are computations? The Computational Lens. Course information and topics. Why Theory? Class Preview. The most fundamental open question of CS: graph coloring ?!? Proof techniques (and an example).
Reading: Chapter 0, Introduction, Sipser
– extra reading: Barak’s chapters 0 and 1
Additional:
What between Ogres, Onions, Parfait, and good proofs?
- Mathematics and Computation, very interesting book by Avi Wigderson, I highly recommend Chapter 20
- An accessible article I wrote of this book (focusing on the computational lens)
- The Computational Lens plus a talk
- Four talks: lens of computation in the sciences
- Panel Discussion
And if my megalomaniac view of the computational lens is not megalomaniac enough, I suggest reading The Last Question by Isaac Asimov
Tuesday 10/4: Deterministic Finite Automata, Closure Properties, Nondeterminism, equivalence of DFSa and NFAs, regular expression and the languages they correspond to.
Assignments: Watch videos 5-11; PSet 1 published
PPTX: 5p-DFA-overview 6p-DFAs 7p-DFAclosure1 8p-NFAs 9p-NFA2DFA 10p-Closure2 11p-RegExp
PDFs: 5p-DFA-overview 6p-DFAs 7p-DFAclosure1 8p-NFAs 9p-NFA2DFA 10p-Closure2 11p-RegExp
Reading (for next week as well): Chapter 1, Sipser
Tuesday 10/11: Non-Regular Languages, The Pumping Lemma, An Algorithm for Minimizing a DFA. The Myhill-Nerode Theorem, Learning DFAs
Assignments: Watch videos 12-15; PSet 2 published; PSet 1 submitted
PPTX: 12p-pumping 13p-minimizingDFA 14p-Myhill-Nerode 15p-Learning-DFA
PDF: 12p-pumping 13p-minimizingDFA 14p-Myhill-Nerode 15p-Learning-DFA
Reading: Chapter 1.4, Sipser; A note on DFA minimization and Myhill-Nerode
Additional:
Tuesday 10/18 Streaming Algorithms, Communication Complexity + Begin Turing Machines (deciding vs. recognizing)
Assignments: Watch videos 16-19; PSet 3 published; PSet 2 submitted
PPTX: 16p-streaming 17p-Communication-Complexity 18p-TM-overview 19p-TMs
PDF: 16p-streaming 17p-Communication-Complexity 18p-TM-overview 19p-TMs
- A note on streaming Algorithms;
- Article on finding frequent elements
- Survey on Communication Complexity
- Turing’s paper.
- Sipser 3
- videos on asymptotic notation (among other preliminaries and relevant videos):
Tuesday 10/25: Continue Turing Machines: , Multitape TM, Universal Turing Machines, Nondeterministic Turing Machines, Undecidable and Unrecognizable, A_TM is unrecognizable, Mapping Reductions
Assignments: Watch videos 20-24; PSet 4 published; PSet 3 submitted
PPTX: 20p-TM-variants 21p-Universal-TM 22p-counting-argument 23p-concrete-undecidable 24p-mapping-reductions
PDF: 20p-TM-variants 21p-Universal-TM 22p-counting-argument 23p-concrete-undecidable 24p-mapping-reductions
- Sipser 4
Tuesday 11/1: Rice’s Theorem, Oracle Machines, Hierarchy of Undecidable Problems, Self Reference, Self Reference, Foundation of Mathematics, Kolmogorov Complexity
Assignments: Watch videos 25-29; PSet 5 published; PSet 4 submitted
PPTX: 25p-rices-theorem 26p-oracle-reductions 27p-self-reference 28p-logic 29p-Kolmogorov-Complexity
PDF: 25p-rices-theorem 26p-oracle-reductions 27p-self-reference 28p-logic 29p-Kolmogorov-Complexity
- Sipser 5, 6
- Note on Rice’s Theorem
- Note on Computability and Logic
- Rosencrantz & Guildenstern Are Dead (1990) – Heads Scene
- Note on Kolmogorov Complexity
Thursday 11/3: midterm
Tuesday 11/8: democracy day Time Complexity, P, Time Hierarchy Theorems, NP and Polynomial-Time (Mapping) Reductions
Assignments: Watch videos 30-33;
PPTX: 30p-Complexity-overview 31p-time-complexity 32p-NP 33p-poly-time-reductions
PDF: 30p-Complexity-overview 31p-time-complexity 32p-NP 33p-poly-time-reductions
- Sipser 7-7.3, 9.1
- Note on NP-Completeness
Tuesday 11/15: NP-Completeness, Cook-Levin Theorem, More NP-Completeness through Poly-Time Reductions, coNP
Assignments: Watch videos 34-36; PSet 6 published; PSet 5 submitted
PPTX: 34p-NPC-and-CL 35p-more-NPC 36p-co-NP
PDF: 34p-NPC-and-CL 35p-more-NPC 36p-co-NP
Tuesday 11/22 thanksgiving
Thursday 11/24 thanksgiving
Tuesday 11/29 Oracles, Polynomial Hierarchy, Space Complexity, advanced topics on proofs (PCPs, Hardness of Approximation, IPs, Zero-Knowledge)
Assignments: Watch videos 37-39; PSet 7 published; PSet 6 submitted
PPTX: 37p-polynomial-hierarchy 38p-space-complexity 39p-Proofs++
PDF: 37p-polynomial-hierarchy 38p-space-complexity 39p-Proofs++
- Finish Sipser 7, Sipser 8.2, 9.1, 9.2
Advanced (not required) reading:
- Arora-Barak Chapters 8 and 18
- Where is Waldo (Applied Kid Cryptography)
Tuesday 12/6: Course Wrap-Up, Computational Lens, Randomness and Pseudorandomness, Algorithmic Fairness
Assignments: Watch videos 40-42; PSet 7 submitted
PPTX: 40p-algorithmic-fairness 41p-Randomness 42p-Parting-Thoughts
PDF: 40p-algorithmic-fairness 41p-Randomness 42p-Parting-Thoughts
Meeting: Ask me Anything
Advanced (not required) reading:
- Arora-Barak Chapters 7, 16
- The Ultimate Question of Life, the Universe, and Everything. What is this “computer of such infinite and subtle complexity that organic life itself formed part of its operational matrix.?” Note that the final video of the class is number 42 (spooky!)
Omer Reingold 