๋ณธ๋ฌธ ๋ฐ”๋กœ๊ฐ€๊ธฐ

๐“ก๐“ธ๐“ธ๐“ถ5: ๐’ฆ๐‘œ๐“‡๐‘’๐’ถ ๐’ฐ๐“ƒ๐’พ๐“‹/Artificial Intelligence(COSE361)

(17)
[์ธ๊ณต์ง€๋Šฅ] 17. Making Sequential Decisions 1. Markov Decision Process 1) Episodic Decisions - nondeterministic, partially observableํ•œ ์ƒํ™ฉ์—์„œ๋Š” ๋‹ค์Œ state๋ฅผ ํ™•์‹ ํ•  ์ˆ˜ ์—†๋‹ค. ๋”ฐ๋ผ์„œ observations e๊ฐ€ ์ฃผ์–ด์กŒ์„ ๋•Œ outcome์€ outcome s'๊ฐ€ ๋งž์„ ๋•Œ์˜ ํ™•๋ฅ ๋กœ ๋‚˜ํƒ€๋‚ผ ์ˆ˜ ์žˆ๋‹ค. P(Result(a) = s' | a, e) - Utility function U(s)๋Š” ๊ฐ state์˜ ์ข‹์€ ์ •๋„๋ฅผ ๋‚˜ํƒ€๋‚ด๋Š” ํ•˜๋‚˜์˜ ์ˆซ์ž์ด๋‹ค. ์ฃผ์–ด์ง„ evidence์—์„œ ์–ด๋–ค action์„ ์ทจํ–ˆ์„ ๋•Œ์˜ expected utility ๊ฐ’์€, ๊ทธ action์„ ์ทจํ–ˆ์„ ๋•Œ ๋‚˜์˜ค๋Š” outcome์˜ average utility value๊ฐ’์ด๋‹ค. EU(a | e) = ∑(s') P(Res..
[์ธ๊ณต์ง€๋Šฅ] 16. Hidden Markov Models (HMM) 1. Definition of Hidden Markov Model : ์‹œ๊ฐ„์˜ ๊ฐœ๋…์ด ํฌํ•จ๋œ ํ™•๋ฅ  ๋ชจ๋ธ - single discrete random variable๋กœ ์ด๋ฃจ์–ด์ ธ ์žˆ๋Š” probabilistic model. - state variable Xt ๋Š” ์ •์ˆ˜ 1, ... , S๋ฅผ ๊ฐ€์งˆ ์ˆ˜ ์žˆ์œผ๋ฉฐ, S๋Š” ๊ฐ€๋Šฅํ•œ states์˜ ์ˆ˜์ด๋‹ค. - transition model P(Xt|Xt-1) ์€ S*S์˜ matrix T์ด๋‹ค. (T_ij = P(Xt = j | X_t-1 = i) - evidence variable Et๋Š” ๊ฐ state์—์„œ specifyํ•˜๊ณ , ๊ฐ state i์—์„œ P(et | Xt = i) ๋ฅผ ํ†ตํ•ด state๊ฐ€ et๋ฅผ ์•ผ๊ธฐํ•˜๋Š”์ง€์— ๋Œ€ํ•ด ์•Œ ์ˆ˜ ์žˆ๋‹ค. ์ด value๋Š” ํŽธ์˜์ƒ S*S์˜ diagona..
[์ธ๊ณต์ง€๋Šฅ] 15. Probabilistic Reasoning over Time (PRoT) 1. Introduction - ์„ธ์ƒ์„ ๋ช‡ ๊ฐœ์˜ observable ํ•˜๊ฑฐ๋‚˜ ๊ทธ๋ ‡์ง€ ์•Š์€ random variables์„ ํฌํ•จํ•˜๊ณ  ์žˆ๋Š” ์–ด๋–ค snapshot ๋˜๋Š” time slices์˜ ์ง‘ํ•ฉ์ด๋ผ๊ณ  ์ƒ๊ฐํ•˜์ž. - Xt : state variables at time t , ์‹œ๊ฐ„์ด t์ผ ๋•Œ์˜ state variable๋“ค์˜ ์ง‘ํ•ฉ. ๊ด€์ธกํ•  ์ˆ˜ ์—†๋Š”, ๊ทธ ๊ฐ’์„ ์•Œ ์ˆ˜ ์—†๋Š” variable์ด๋‹ค. - Et : evidence variables, ๊ด€์ธกํ•  ์ˆ˜ ์žˆ๋Š”, ์•Œ ์ˆ˜ ์žˆ๋Š” variable - transition model : state์ด ์–ด๋–ป๊ฒŒ ๋ณ€ํ•˜๋Š”์ง€ ์ •์˜ํ•œ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด t-1 ์ผ ๋•Œ state์ด X์ผ ๋•Œ, ๋‹ค์Œ state์ด X ์ผ ํ™•๋ฅ  - sensor model : evidence variable์ด ๊ฐ’์ด ์–ด๋–ป๊ฒŒ ..
[์ธ๊ณต์ง€๋Šฅ] 14. Bayesian Networks 1. Definition of Bayesian Networks 1) Bayesian networks (= belief networks, Bayesian belief networks, graphical models) - Directed graph์ž„ - Nodes (= Random variable)์™€ Links ๋กœ ๊ตฌ์„ฑ๋จ - node X๋กœ๋ถ€ํ„ฐ node Y๋กœ์˜ links๋กœ ์ด๋ฃจ์–ด์ง„ directed acyclic graph(DAG) - X๋Š” causes, Y๋Š” effects - Conditional probability distribution P(Xi | Parents(Xi) ) : Xi ์˜ ๋ถ€๋ชจ ๋…ธ๋“œ๋“ค์ด ๋ฐœ์ƒํ•  ๋•Œ Xi๊ฐ€ ๋ฐœ์ƒํ•  ํ™•๋ฅ  ์ฆ‰ ~๊ฐ€ ๋ฐœ์ƒํ•  ๋•Œ์˜ ํ™•๋ฅ ์„ ํŠธ๋ฆฌ์ฒ˜๋Ÿผ..? ๊ทธ๋ž˜ํ”„๋กœ ๋‚˜ํƒ€๋ƒ„ 2) Condit..
[์ธ๊ณต์ง€๋Šฅ] 13. Probability and Statistics 1. Review of Probabilities and Statics 1) Elements of Probability - Random experiment : ์‹œํ–‰, Nondeterministic - Sample space : ์ „์ฒด์ง‘ํ•ฉ ์š”์†Œ๋“ค. Mutually exclusive(๋ฐฐ๋ฐ˜์‚ฌ๊ฑด)์ด๊ณ  exhaustive(ํ•ฉ์ง‘ํ•ฉ = ์ „์ฒด์ง‘ํ•ฉ) ์ด๋‹ค. - Events : ์‚ฌ๊ฑด, Sample space์˜ ๋ถ€๋ถ„์ง‘ํ•ฉ - Probability : ํ™•๋ฅ . the proportion, or relative frequency (๋น„์œจ, ์ƒ๋Œ€์  ๋นˆ๋„์ˆ˜) / degree of belief (๊ธฐ๋Œ“๊ฐ’) 2) Axioms of Probability Theory - ํ™•๋ฅ ์€ 0๊ณผ 1 ์‚ฌ์ด์ž„ - Sample space์˜ ๋ชจ๋“  ํ™•๋ฅ ์„ ๋”ํ•˜๋ฉด 1 -..
[์ธ๊ณต์ง€๋Šฅ] 8~9. First-Order Logic(FOL) 1. Syntax and Semantics : propositional logic, ๋ช…์ œ๋…ผ๋ฆฌํ•™์€ ์ด ์„ธ์ƒ์˜ ๋ชจ๋“  ๋ฌธ์ œ๋ฅผ propositional symbol๋กœ ๋‚˜ํƒ€๋‚ด๊ณ  ์˜ค์ง true/false์˜ ๊ฐ’๋งŒ ๊ฐ€์ง„๋‹ค. ๋”ฐ๋ผ์„œ ์ด ์„ธ์ƒ์— ๋งŽ์€ object๋“ค์„ ํ‘œํ˜„ํ•˜๊ธฐ์—” ๋ถ€์กฑํ•œ ๋ฉด์ด ์žˆ๋‹ค. (lack the expressive power) ์šฐ๋ฆฌ๊ฐ€ ์ž์—ฐ์–ด์˜ ๋ฌธ๋ฒ•์„ ์‚ดํŽด๋ณด๋ฉด, ๋Œ€๋ถ€๋ถ„์˜ elements๋“ค์ด ๋ช…์‚ฌ์™€ ๋ช…์‚ฌ๊ตฌ๋กœ object๋ฅผ ๋‚˜ํƒ€๋‚ด๊ณ , ๋™์‚ฌ๋‚˜ ๋™์‚ฌ๊ตฌ๋กœ object๊ฐ„์˜ relation์„ ๋‚˜ํƒ€๋‚ธ๋‹ค. ๊ทธ๋ฆฌ๊ณ  ์ด relation์ค‘์˜ ์ผ๋ถ€๋Š” ์–ด๋–ค input์ด ์ฃผ์–ด์กŒ์„ ๋•Œ ํ•˜๋‚˜์˜ value๋ฅผ ๋‚ด๋ฑ‰๋Š” function(object์˜ 1:1 ๊ด€๊ณ„)์ด๊ธฐ๋„ ํ•˜๋‹ค. First-order logic์€ ์ด ์„ธ์ƒ์—” object์™€ ์ด๋“ค์„ ..
[์ธ๊ณต์ง€๋Šฅ] 7. Propositional Logic - 3 5) Analysis of Resolution Algorithm resolution rule์€ ์–ด๋–ค completeํ•œ search algorithm๊ณผ ๊ฒฐํ•ฉํ•  ๋•Œ completeํ•œ inference algorithm์„ ๋งŒ๋“ค์–ด๋‚ธ๋‹ค. PL-Resolution์ด completeํ•จ์„ ๋ณด์ด๊ธฐ ์œ„ํ•ด์„œ resolution closure = RC(S)๋ฅผ ์ •์˜ํ•œ๋‹ค. RC(S)๋ž€ "clause์˜ ์ง‘ํ•ฉ S์— ๋“ค์–ด์žˆ๋Š” clause์™€ ๊ทธ clause๋ผ๋ฆฌ resolution์„ ํ†ตํ•ด ๋งŒ๋“ค์–ด์ง„ ๋ชจ๋“  clause๋“ค"์— ๋˜ resolution์„ ๋ฐ˜๋ณตํ•ด์„œ ์ ์šฉํ•ด์„œ ๋งŒ๋“ค ์ˆ˜ ์žˆ๋Š” ๋ชจ๋“  clause๋“ค์˜ ์ง‘ํ•ฉ์„ ๋งํ•œ๋‹ค. RC(S)๊ฐ€ ์œ ํ•œํ•˜๋‹ค๋Š” ๊ฒƒ์€ ๋ช…๋ฐฑํžˆ ์•Œ ์ˆ˜ ์žˆ๋‹ค. ์œ ํ•œํ•œ symbol๋“ค๋กœ ๋งŒ๋“ค ์ˆ˜ ์žˆ๋Š” clause๋Š” ์œ ํ•œ๊ฐœ ์กด์žฌํ•  ๊ฒƒ์ด๊ณ , ..
[์ธ๊ณต์ง€๋Šฅ] 7. Propositional logic - 2 3. Theorem Proving : Theorem Proving์ด๋ž€ ์šฐ๋ฆฌ๊ฐ€ ์ด๋ฏธ ์•Œ๊ณ ์žˆ๋Š” ์‚ฌ์‹ค(Knowledge base)์˜ sentence๋“ค์— ์ถ”๋ก  ๊ทœ์น™(Inference rule)์„ ์ ์šฉํ•ด์„œ ์ƒˆ๋กœ์šด ์‚ฌ์‹ค์„ ์•Œ์•„๋‚ด๋Š” ๊ฒƒ/ ์ƒˆ๋กœ์šด ๋ฌธ์žฅ์„ ๋งŒ๋“ค์–ด๋‚ด๋Š” ๊ฒƒ์„ ๋งํ•œ๋‹ค. ๋งŒ์•ฝ model์˜ ๊ฐœ์ˆ˜๊ฐ€ ๋งŽ๊ณ , proof์˜ ๊ธธ์ด๊ฐ€ ์งง๋‹ค๋ฉด, theorem proving์ด model checking๋ณด๋‹ค ๋” ํšจ์œจ์ ์ด๋‹ค. 1) Inference Rules - Modus Ponens : α⇒β์ด๊ณ  α๊ฐ€ true๋ฉด β๋„ true๋‹ค. - And-Elimination : α and β๊ฐ€ true์ด๋ฉด α๋Š” ํ•ญ์ƒ true์ด๋‹ค - Logical equivalences : α and β๊ฐ€ true์ด๋ฉด β and α๋„ true์ด๋‹ค. - α ์™€..
[์ธ๊ณต์ง€๋Šฅ] 7. Propositional logic - 1 1. Propositional Logic (๋ช…์ œ ๋…ผ๋ฆฌํ•™) 1) Example ์ด๋ฒˆ ๋‹จ์› ๋‚ด๋‚ด ์ง€๊ฒน๋„๋ก ๋ณด๊ฒŒ ๋  Wumpus World game์ด๋‹ค. ์ด ๊ฒŒ์ž„์€ (1,1)์—์„œ ์šฉ์‚ฌ๊ฐ€ ์ถœ๋ฐœํ•ด ๊ดด๋ฌผ๊ณผ ํ•จ์ •์„ ํ”ผํ•ด gold๋ฅผ ๋ฌด์‚ฌํžˆ ์ฐพ๋Š” ๊ฒƒ์ด ๋ชฉํ‘œ์ด๋‹ค. ํ•จ์ •์ด ์žˆ์œผ๋ฉด ํ•จ์ •์˜ ์‚ฌ๋ฐฉ์œผ๋กœ breeze ๋ฐ”๋žŒ์ด ๋ถˆ๊ณ , ๊ดด๋ฌผ์ด ์žˆ์œผ๋ฉด ๊ดด๋ฌผ์˜ ์‚ฌ๋ฐฉ์œผ๋กœ ์•…์ทจ strench๊ฐ€ ํ’๊ธด๋‹ค. ์ด ํžŒํŠธ๋ฅผ ์ด์šฉํ•ด ์ตœ์†Œํ•œ์œผ๋กœ ์›€์ง์—ฌ ๋ฌด์‚ฌํžˆ gold๋ฅผ ์ฐพ์œผ๋ฉด ๋œ๋‹ค. ์šฉ์‚ฌ๊ฐ€ ์–ด๋–ค Action์„ ์ทจํ•  ๋•Œ๋งˆ๋‹ค cost๊ฐ€ ๋“ ๋‹ค. ์ด๊ฑด ์•Œ์•„๋„ ๋˜๊ณ  ๋ชฐ๋ผ๋„ ๋˜๋Š”๋ฐ ์šฉ์‚ฌ๋Š” ํ™”์‚ด์„ ์  ์ˆ˜ ์žˆ๋‹ค. ํ™”์‚ด์„ ์˜๋ฉด ์ง์„  ๋ฐฉํ–ฅ์œผ๋กœ ์ญ‰ ๋‚ ์•„๊ฐ€๋Š”๋ฐ, ๊ดด๋ฌผ์ด ๋งž์œผ๋ฉด ์†Œ๋ฆฌ๋ฅผ ์ง€๋ฅด๋ฉด์„œ ์ฃฝ๋Š”๋‹ค. ์ด ์†Œ๋ฆฌ๋ฅผ ๋“ฃ๊ณ  ๊ดด๋ฌผ์˜ ์œ„์น˜๋ฅผ ์˜ˆ์ธกํ•  ์ˆ˜ ์žˆ๋‹ค. ์šฉ์‚ฌ๋Š” ์ž๊ธฐ๊ฐ€ ์žˆ๋Š” ์นธ์—์„œ ์ •๋ณด..
[์ธ๊ณต์ง€๋Šฅ] 6. Constraint Satisfaction Problems 1. Problem Formulation as a CSP : ๋ฌธ์ œ ์ƒํ™ฉ์„ ํ‘œํ˜„ํ•˜๋Š” ๋ฐฉ๋ฒ• ์ค‘์—์„œ๋Š” factored represetation์ด ์žˆ๋‹ค. ์–ด๋–ค ๋ฌธ์ œ๊ฐ€ ์–ด๋–ค ์ œํ•œ์‚ฌํ•ญ, ์กฐ๊ฑด๋“ค(Constraint)๋ฅผ ๋งŒ์กฑ์‹œ์ผœ์•ผ ํ’€๋ฆฐ๋‹ค๊ณ  ํ•ด๋ณด์ž. Factored representation์€ ๊ฐ states๋ฅผ variable์„ ํ™œ์šฉํ•ด ํ‘œํ˜„ํ•˜๋Š” ๊ฒƒ์ด๊ณ (goal state ๋˜ํ•œ ๊ทธ๋ ‡๋‹ค), ์ด variable์— ํ• ๋‹น๋œ value๊ฐ’์ด constraint์„ ๋งŒ์กฑํ•˜๋Š” ๊ฐ’์ผ ๋•Œ ๋ฌธ์ œ๋ฅผ ํ•ด๊ฒฐํ–ˆ๋‹ค๊ณ  ๋งํ•˜๋Š” ๊ฒƒ์ด๋‹ค. ์ฆ‰, ์–ด๋–ค ๊ฐ’์„ ํ• ๋‹นํ•  ์ˆ˜ ์žˆ๋Š” variable๊ณผ ๋…ผ๋ฆฌ์—ฐ์‚ฐ์ž๋ฅผ ์ด์šฉํ•ด ์–ด๋–ค ๋ฌธ์ œ ์ƒํ™ฉ์„ ํ‘œํ˜„ํ•œ๋‹ค๊ณ  ํ•  ๋•Œ, ์ด๋Ÿฐ ๋ฌธ์ œ๋“ค์„ Constraint Satisfaction Problem์ด๋ผ๊ณ  ํ•œ๋‹ค. ( ์ง€๊ธˆ๊นŒ์ง€๋Š” autonomic..