Skip to content

CSCI3230

Fundamentals of Artificial Intelligence

Search and adversarial games, logical and rule-based inference, planning, probabilistic inference and Bayesian networks, sequential decisions, and learning with decision trees and neural networks.

30 chapters · 31 sectionsEN · 繁 · 简
My course notes

Your progress

0/ 31 sections marked complete

Next section: 1.1 Heuristic search and adversarial games

Completion marks are saved in this browser.

This course in the web

In the CSCI3230 notes

Theorem cards
68
Definitions
95
Counterexamples
33
Concept lenses
26
Histories of ideas
1
Proof X-rays
36

Sections with a history of an idea

None of these is on the timeline yet.

No concept lens from this course is recorded in the web yet.

Open this course in the knowledge map

Counts cover the 31 public sections of this course and what is written in them, not recorded connections. 0 of these sections are indexed into the web in detail so far.

In the webTheorem cards 68 · Definitions 95 · Counterexamples 33 · Concept lenses 26 · Histories of ideas 1 · Proof X-rays 36

None of these is on the timeline yet.

No concept lens from this course is recorded in the web yet.

Open this course in the knowledge map

Counts cover the 31 public sections of this course and what is written in them, not recorded connections. 0 of these sections are indexed into the web in detail so far.

Chapter 2 · 1 section

Knowledge and planning

Separate truth from derivability, prove Horn-rule inference sound, and verify state-changing plans.

Theorems 2 · Definitions 4Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 2.1

    Logical inference and STRIPS planning

    Separate truth from derivability, prove Horn-rule inference sound, and verify state-changing plans.

Chapter 3 · 1 section

Learning and representation

Derive perceptron and neural-network updates, prove a separable-data mistake bound, and inspect decision-tree information gain.

Theorems 3 · Definitions 4Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 3.1

    Learning rules and the limits of representation

    Derive perceptron and neural-network updates, prove a separable-data mistake bound, and inspect decision-tree information gain.

Chapter 4 · 1 section

First-order inference

Construct most general unifiers, justify resolution by models, and prove finite propositional refutation completeness.

Theorems 2 · Definitions 3Counterexamples 1Lenses 1Proof X-rays 2

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 4.1

    Unification, resolution, and exact inference limits

    Construct most general unifiers, justify resolution by models, and prove finite propositional refutation completeness.

Chapter 5 · 1 section

Structured knowledge and expert systems

Give finite semantics to inheritance and event scripts, then construct terminating goal-directed rule evaluation with explicit explanations.

Theorems 2 · Definitions 5Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 5.1

    Frames, scripts, and explainable expert reasoning

    Give finite semantics to inheritance and event scripts, then construct terminating goal-directed rule evaluation with explicit explanations.

Chapter 6 · 1 section

Plan-space reasoning

Give complete finite planning procedures, prove causal-link plan validity, and expose threats and persistence assumptions.

Theorems 2 · Definitions 4Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 6.1

    State-space search and protected causal links

    Give complete finite planning procedures, prove causal-link plan validity, and expose threats and persistence assumptions.

Chapter 7 · 1 section

Mathematical language models

Prove CKY parsing and derive HMM forward and Viterbi recurrences, then estimate parameters from labeled sequences.

Theorems 3 · Definitions 3Counterexamples 1Lenses 1Proof X-rays 2

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 7.1

    Grammars, ambiguity, and hidden-state language models

    Prove CKY parsing and derive HMM forward and Viterbi recurrences, then estimate parameters from labeled sequences.

Chapter 8 · 1 section

Vision foundations

Derive normalized Sobel responses and their noise variance, examine localization after smoothing, and prove threshold-connectivity extraction.

Theorems 2 · Definitions 5Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 8.1

    Image gradients, noise, and edge connectivity

    Derive normalized Sobel responses and their noise variance, examine localization after smoothing, and prove threshold-connectivity extraction.

Chapter 9 · 1 section

Probabilistic inference

Conditional independence and finite posterior calculations.

Theorems 1 · Definitions 2Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 9.1

    Conditional independence and evidence

    Compute a finite Bayesian posterior and identify exactly where conditional independence is used.

Chapter 10 · 1 section

Sequential decisions

Discounted stochastic planning and optimality.

Theorems 4 · Definitions 2Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 10.1

    Discounted decisions and value iteration

    Prove Bellman contraction, certify value error and derive an optimal stationary policy from the exact fixed point.

Chapter 11 · 1 section

HMM smoothing

State and transition posteriors from forward–backward messages.

Theorems 1 · Definitions 1Counterexamples 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 11.1

    Forward–backward smoothing in a hidden Markov model

    Derive state and transition posteriors from prefix and suffix probabilities.

Chapter 12 · 1 section

Factor-based inference

Exact inference using finite factor tables.

Theorems 2 · Definitions 3Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 12.1

    Exact inference by variable elimination

    Derive sum-product variable elimination, prove its invariant and compute a posterior with explicit evidence.

Chapter 17 · 1 section

Truth maintenance

Recompute grounded conclusions after assumptions are withdrawn, and see why local support counts can fail on cycles.

Theorems 1 · Definitions 2Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 17.1

    Truth maintenance under retraction

    Recompute grounded conclusions after assumptions are withdrawn, and see why local support counts can fail on cycles.

Chapter 18 · 1 section

Planning and relaxation

Develop positive-goal regression and admissible delete-relaxed planning bounds, with a worked delivery and return example.

Theorems 5 · Definitions 1Counterexamples 1Proof X-rays 2

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 18.1

    STRIPS planning and delete-relaxed bounds

    Develop positive-goal regression and admissible delete-relaxed planning bounds, with a worked delivery and return example.

Chapter 20 · 1 section

Rule-based inference

Forward chaining, least models and explicit worklist costs.

Theorems 1 · Definitions 3Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 20.1

    Horn rules, forward chaining and countermodels

    Trace a finite rule system and connect least-model proofs to a worklist algorithm.

Chapter 21 · 1 section

Objectives and agent models

Specify an agent's objective, derive action-dependent Bayesian state estimation and finite-horizon control, and distinguish information, architecture and model error.

Theorems 2 · Definitions 2Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 21.1

    Rational objectives and finite belief-state control

    Specify an agent's objective, derive action-dependent Bayesian state estimation and finite-horizon control, and distinguish information, architecture and model error.

Chapter 23 · 1 section

Simulated annealing

Derive the Metropolis acceptance rule, trace a search, and separate a fixed-temperature equilibrium from a cooling guarantee.

Theorems 2 · Definitions 2Counterexamples 2Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 23.1

    Simulated annealing on a finite search space

    Derive the Metropolis acceptance rule, trace a search, and separate a fixed-temperature equilibrium from a cooling guarantee.

Chapter 25 · 1 section

Rule execution policies

Give a finite rule engine explicit firing, conflict-resolution and termination semantics, and distinguish its execution trace from logical closure.

Theorems 1 · Definitions 3Counterexamples 2Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 25.1

    Rule agendas, conflicts and state changes

    Give a finite rule engine explicit firing, conflict-resolution and termination semantics, and distinguish its execution trace from logical closure.

Chapter 26 · 1 section

Typed expert policies

Build a finite typed expert-policy tool, distinguish derivation from action selection, and prove when alternative evidence supports survive withdrawal.

Theorems 1 · Definitions 4Counterexamples 1Lenses 1History 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 26.1

    Typed evidence, acquired policies and rule provenance

    Build a finite typed expert-policy tool, distinguish derivation from action selection, and prove when alternative evidence supports survive withdrawal.

Chapter 27 · 1 section

Acquiring rule tools

Account for entropy reductions through a finite tree, prove tree-to-rule equivalence, and audit an executable Boolean diagnostic tool.

Theorems 2 · Definitions 4Counterexamples 1Lenses 1Proof X-rays 2

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 27.1

    Acquiring decision trees and compiling rule tools

    Account for entropy reductions through a finite tree, prove tree-to-rule equivalence, and audit an executable Boolean diagnostic tool.

Chapter 28 · 1 section

Decision-tree pruning

Trace reduced-error pruning and distinguish its validation guarantee from generalization.

Theorems 1 · Definitions 3Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 28.1

    Decision-tree pruning with validation data

    Trace reduced-error pruning and distinguish its validation guarantee from generalization.

Chapter 29 · 1 section

Evaluating supervised learning

Train and compare two finite classifiers without leaking evaluation data, calculate confusion matrices, and prove a validation-selection bound.

Theorems 1 · Definitions 3Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 29.1

    A controlled classification experiment

    Train and compare two finite classifiers without leaking evaluation data, calculate confusion matrices, and prove a validation-selection bound.

Chapter 30 · 1 section

Computational graphs

Prove reverse-mode differentiation on a finite graph, derive softmax cross-entropy gradients, and follow complete simultaneous parameter updates.

Theorems 2 · Definitions 4Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 30.1

    Computational graphs and softmax learning

    Prove reverse-mode differentiation on a finite graph, derive softmax cross-entropy gradients, and follow complete simultaneous parameter updates.

Chapter 31 · 1 section

Shared neural operators

Derive translation-equivariant shared filters and finite recurrent backpropagation, following small applications and exact training steps.

Theorems 2 · Definitions 3Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 31.1

    Shared convolutions and recurrent gradients

    Derive translation-equivariant shared filters and finite recurrent backpropagation, following small applications and exact training steps.

Chapter 32 · 1 section

Probabilistic parsing: sums, maxima and ambiguity

Derive inside and Viterbi chart recurrences for a finite probabilistic grammar and interpret their different answers.

Theorems 3 · Definitions 2Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 32.1

    Probabilistic parsing: sums, maxima and ambiguity

    Derive inside and Viterbi chart recurrences for a finite probabilistic grammar and interpret their different answers.

Chapter 33 · 1 section

Bayesian networks

Define a network by local conditional tables, prove its factorization and read conditional independence from the graph by d-separation.

Theorems 2 · Definitions 3Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 33.1

    Bayesian networks: factorization and d-separation

    Define a Bayesian network by local conditional tables, prove that its product is a joint distribution, and read conditional independence from the graph.

Chapter 34 · 1 section

Policy iteration

Evaluate stationary policies exactly, prove policy improvement, and show termination at an optimal policy.

Theorems 4 · Definitions 2Counterexamples 1Lenses 1Proof X-rays 1

Chapter outline →See this chapter on the web →Not yet indexed in the web

  1. 34.1

    Policy iteration for discounted decision problems

    Evaluate a stationary policy by a linear system, prove the policy improvement theorem, and show that policy iteration terminates at an optimal policy.

Vercel visit statistics and public-page performance measurements each require separate consent. Neither is needed to use the website.

Website function storage (used when needed)

Security, sign-in and your chosen preferences use storage when needed. This is separate from optional statistics.

Withdrawing stops future statistics. The page reloads if analytics code needs unloading. It does not delete data already sent.

Read Cookie and storage details