Optimisation

When classical beats quantum, and how to read an optimisation claim.

1 lesson, 33 course topics, about 192 min at Plain. Opened pages are remembered in this browser only.

Start here

Can a classical machine out-optimise a quantum one?about 3 min at Plain

The lessons, in reading order

  1. Can a classical machine out-optimise a quantum one?

    You'll be able to say when a classical or analog machine beats a quantum one on an optimisation problem.

    about 3 min at Plain Plain, Working or Formal

The course: The Feasible Region

Operations research, from linear programming to game theory and reinforcement learning. Open the course page.

Foundations Regions, corners and curves5 topics, about 26 min at Plain
  1. The feasible region

    about 4 min at Plain One depth After: Can a classical machine out-optimise a quantum one?

  2. Linear programming and the vertex theorem

    about 4 min at Plain One depth After: The feasible region

  3. The simplex algorithm

    about 6 min at Plain One depth After: Linear programming and the vertex theorem

  4. Duality and shadow prices

    about 5 min at Plain One depth After: The simplex algorithm

  5. Quadratic & semidefinite programming

    about 7 min at Plain One depth After: Duality and shadow prices

Discrete When the answer must be whole, and when exact methods run out5 topics, about 33 min at Plain
  1. Integer programming

    about 5 min at Plain One depth After: Quadratic & semidefinite programming

  2. Branch and bound

    about 5 min at Plain One depth After: Integer programming

  3. Constraint satisfaction

    about 7 min at Plain One depth After: Branch and bound

  4. Greedy algorithms & approximation guarantees

    about 8 min at Plain One depth After: Constraint satisfaction

  5. Metaheuristics

    about 8 min at Plain One depth After: Greedy algorithms & approximation guarantees

Networks Flows, assignments and routes4 topics, about 19 min at Plain
  1. Network flow and min-cut

    about 5 min at Plain One depth After: Metaheuristics

  2. The transportation problem

    about 6 min at Plain One depth After: Network flow and min-cut

  3. Matching and assignment

    about 4 min at Plain One depth After: The transportation problem

  4. Shortest paths and dynamic programming

    about 4 min at Plain One depth After: Matching and assignment

Uncertainty Deciding when the world is not certain3 topics, about 19 min at Plain
  1. Markov decision processes & reinforcement learning

    about 8 min at Plain One depth After: Shortest paths and dynamic programming

  2. Stochastic & robust optimization

    about 7 min at Plain One depth After: Markov decision processes & reinforcement learning

  3. Queueing theory

    about 4 min at Plain One depth After: Stochastic & robust optimization

Many goals and players When more than one objective, or more than one decision-maker, is involved6 topics, about 31 min at Plain
  1. Multi-objective optimization

    about 6 min at Plain One depth After: Queueing theory

  2. Nash equilibrium

    about 5 min at Plain One depth After: Multi-objective optimization

  3. Zero-sum games and minimax

    about 5 min at Plain One depth After: Nash equilibrium

  4. Cooperative games and the Shapley value

    about 5 min at Plain One depth After: Zero-sum games and minimax

  5. Mechanism design and auctions

    about 5 min at Plain One depth After: Cooperative games and the Shapley value

  6. Evolutionary game theory

    about 5 min at Plain One depth After: Mechanism design and auctions

Games, part two Time, secrets, repetition, learning, signals, traffic, hardness and a quantum game9 topics, about 55 min at Plain
  1. Games over time: threats and backward induction

    about 5 min at Plain One depth After: Evolutionary game theory

  2. Private information: auctions and Bayesian games

    about 5 min at Plain One depth After: Games over time: threats and backward induction

  3. Repeated games: cooperation that lasts

    about 6 min at Plain One depth After: Private information: auctions and Bayesian games

  4. Learning to play: regret and the minimax value

    about 6 min at Plain One depth After: Repeated games: cooperation that lasts

  5. Correlated equilibrium: a traffic light

    about 6 min at Plain One depth After: Learning to play: regret and the minimax value

  6. What selfishness costs: the price of anarchy

    about 6 min at Plain One depth After: Correlated equilibrium: a traffic light

  7. How hard is an equilibrium?

    about 6 min at Plain One depth After: What selfishness costs: the price of anarchy

  8. A game where physics changes the answer: CHSH

    about 7 min at Plain One depth After: How hard is an equilibrium?

  9. Design a small market and break it

    about 8 min at Plain One depth After: A game where physics changes the answer: CHSH

Capstone Inventory and supply chain1 topic, about 6 min at Plain
  1. Inventory & supply chain

    about 6 min at Plain One depth After: Design a small market and break it

How this track counts toward your rank142 points in all

Your rank is a code distance, d3 up to d25. Points come from questions you answer correctly and game levels you finish, never from opening a page, and each point belongs to one track. Your overall rank is set by your weakest track, and the top rank needs 80% of every track. This track holds 142 points: 1 lesson check question, 1 point each (1); 33 course topic checks, 2 points each (66); 25 game levels, up to 3 points each for gold (75). Its games: Max-Cut (9), The Annealing Volcano (9), Heuristic Arena (6), Be FunSearch (1).

Points needed on this track, when it is your weakest, for each rank
RankPoints
d30
d511
d721
d931
d1142
d1352
d1562
d1773
d1983
d2193
d23104
d25114
Prove itThe Bench: three methods race on one graph Judge a claimThe Ledger: did they do what they said? Keep goingNext track: Post-quantum security