ComputationPath

ComputationPath covers CS theory adjacent to machine learning: what can be computed at all, what can be computed efficiently, what languages and grammars classify, and how proofs are represented and checked. Its audience is readers who want enough computer-science theory to reason clearly about ML without conflating distinct frontiers.

The site is not a complete CS curriculum. It is the focused shelf of CS-theory pages that earn their own home outside the main TheoremPath ML curriculum. Topics like computability and P vs NP are useful background for ML researchers; treating them as core ML prerequisites would misrepresent both fields.

What you will find here

• Computability: Turing machines, decidability, the halting problem, Gödel's incompleteness theorems.
• Complexity: P, NP, NP-completeness, SAT and SMT solvers, and the practical reach of automated reasoning.
• Automata and formal languages: regular languages, context-free grammars, the Chomsky hierarchy, and the machines that recognize them.
• Proofs and types: lambda calculus, dependent type theory, proof assistants like Lean and Coq, and formal verification.

What you will not find here

• A general-purpose introduction to programming, software engineering, or systems design.
• ML methods, optimization theory, or statistical learning foundations. Those live on TheoremPath.
• Claims that any open problem (P vs NP, the Riemann hypothesis, the Collatz conjecture) is on the verge of being resolved.

Author

ComputationPath is written by Robby Sneiderman as part of the Path Network. The companion site TheoremPath covers ML theory, statistics, optimization, and mathematics. Linguistics and language structure live on LinguisticsPath, and philosophy of language and epistemology live on PhilosophyPath.

Contact and corrections

Corrections are welcome. If you find a factual error, a broken link, or a source that should be cited, use the shared contact form.

See also: disclaimer · privacy · terms.