Knowledge Representation and Deduction First Order Logic - PowerPoint PPT Presentation

1 / 14
About This Presentation
Title:

Knowledge Representation and Deduction First Order Logic

Description:

Knowledge Representation and Deduction. First Order ... 2. Review. 3. Review. Contoh penggunaan propositional logic pd wumpus world : Knowledge Base : Awal : ... – PowerPoint PPT presentation

Number of Views:223
Avg rating:3.0/5.0
Slides: 15
Provided by: Micr130
Category:

less

Transcript and Presenter's Notes

Title: Knowledge Representation and Deduction First Order Logic


1
Knowledge Representation and Deduction First
Order Logic
  • Ref Artificial Intelligence A Modern Approach
    ch. 7, 9
  • Rabu, 27 Feb 2002

2
Review
3
Review
  • Contoh penggunaan propositional logic pd wumpus
    world
  • Knowledge Base
  • Awal
  • S1,1 ? W1,1 /\ W1,2 /\ W2,1
  • S2,1 ? W1,1 /\ W2,1 /\ W3,1 /\ W2,2
  • S1,2 ? W1,1 /\ W1,2 /\ W1,3 /\ W2,2
  • S2,2 ? W2,1 /\ W1,2 /\ W2,2 /\ W3,2 /\
    W2,3
  • B1,1 ? P1,1 /\ P1,2 /\ P2,1
  • B2,1 ? P1,1 /\ P2,1 /\ P3,1 /\ P2,2
  • B1,2 ? P1,1 /\ P1,2 /\ P1,3 /\ P2,2
  • B2,2 ? P2,1 /\ P1,2 /\ P2,2 /\ P3,2 /\
    P2,3
  • Sambil jalan
  • S1,1 B1,1
  • Inference
  • Modus ponens W1,1 /\ W1,2 /\ W2,1
  • P1,1 /\ P1,2 /\ P2,1

4
First Order Logic
  • Syntax
  • Constants.
  • Merepresentasikan objek (cth. saya, kamu)
  • Predicates.
  • Merepresentasikan relasi (cth. teman)
  • Functions.
  • Relasi dimana suatu objek tepat berelasi
    (dipetakan ke) dng. 1 objek lain.
  • Variables (cth. x, y, z).
  • Connectives (cth. , /\, \/, ,
    )
  • Equality ().
  • Quantifiers ( ).
  • Term constant variable function(term1,
    term2, ..., termn).
  • Atomic sentence predicate(term1, ..., termn)
    term1 term2.
  • Complex sentence dr. 1/lbh. atomic sentences
    dng. connectives.

5
First Order Logic
  • Semantics
  • Sentences true/false tergantung model
    interpretasi.
  • Model berisi objek relasi antar objek.
  • Interpretasi
  • Constant symbols ? objek.
  • Predicate symbols ? relasi.
  • Function symbols ? functional relations.
  • Atomic sentence predicate(term1, ..., termn)
    true/false
  • True jkk objek yg. direpresentasikan oleh term1,
    ..., termn berada dalam relasi yg.
    direpresentasikan oleh predicate.
  • Contoh
  • Semua orang perlu tidur.
  • x Orang(x) ? y Tidur(y) /\ Perlu(x,y).
  • Ani makan semua yang Budi makan
  • x makan(Budi, x) ? makan(Ani, x).

6
First Order Logic
  • Reasoning
  • Inference rule utk. propositional logic.
  • Substitution subst(x/saya, y/kampus,
    pergi(x,y)) pergi(saya, kampus).
  • Universal Elimination
  • v, ?
  • subst(v/g, ?)
  • Existential Elimination
  • v, ?
  • subst (v/k, ?)
  • Existential Introduction.
  • ?
  • v subst(g/v, ?)

7
Proofing (Primitive Inference Rules)
  • Contoh
  • A angk.1998.
  • B angk.2000.
  • Angk.1998 lebih tua daripada angk.2000.
  • Buktikan A lebih tua dari B.
  • Bukti menggunakan and-introduction, universal
    elimination, modus ponens.
  • Dng. search
  • Operator inference rules.
  • States set of sentences.
  • Goal test cek apakah telah terbukti true /
    false.
  • Kekurangan branching factor bsr.
  • Atasi Cari substitusi yg. simana premise match
    dng. fakta.

8
Unification - Generalized Modus Ponens
  • Unification
  • Unify menghasilkan substitution yg. sesuai dari
    2 atomic sentence.
  • Contoh
  • Unify(knows(John, x), knows(John, Jane))
    x/Jane
  • Unify(knows(John, x), knows(y, mother(y))
    x/mother(John), y/John.
  • Generalized Modus Ponens
  • Atomic sentences pi, pi, q di mana terdpt
    subst(t, pi) subst(t, pi) utk. setiap i
    p1, p2, , pn, (p1 /\ p2 /\ /\ pn ? q)
  • subst(t, q)
  • Contoh
  • p1lbh. tua(A,B), p2lbh.tua(B, C),
  • x, y, z lbh.tua(x,y) /\ lbh.tua(y,z) ?
    lbh.tua(x,z)
  • lbh.tua(A, C).

9
Proofing (Forward Chaining)
  • Fakta baru
  • Add ke KB
  • For tiap rule di mana p unifies dng. suatu
    premis
  • Jk. premis lain telah diketahui
  • Add konklusi ke KB, lanjutkan chaining.

10
Proofing (Backward Chaining)
  • Mulai dari kesimpulan yg. hendak dibuktikan.

11
Proofing (Resolution)
  • Menggunakan 1 inference rule resolution rule.
  • Mirip dng. resolution pd. propositional logic.
  • Generalized Resolution
  • Utk. atom pi, qi, ri, si dimana Unify(pj, qj) t
  • p1 /\ /\ pj /\ pn1 --gt r1 \/ \/ rn2
  • s1 /\ /\ sn3 --gt q1 \/ \/ qk \/
    \/ qn4
  • subst(t, (p1 /\ /\ pj-1 /\ pi1 /\ /\ pn1 /\
    s1 /\ /\ sn3 --gt r1 \/ \/ rn2 \/ q1 \/ \/
    qk-1 \/ qk1 \/ \/ qn4)
  • Contoh lihat papan.

12
Proofing (Resolution with Refutation)
  • Sama dng. resolution, hanya saja kontradiksi dari
    sentence yg. akan dicari juga dimasukkan ke dalam
    KB.
  • Pembuktian by contradiction.
  • Contoh lihat papan.

13
Mengubah FOLMenjadi Clause Form
  • Eliminasi biimplikasi
  • Eliminasi implikasi
  • Reduce scope
  • Standardize variables
  • Skolemization
  • Hilangkan universal quantifier
  • Buat jadi CNF
  • Buat jadi clauses
  • Standardize variables lagi

14
Aplikasi Financial Advisor
  • Ref.Artificial Intelligence Structures and
    Strategies for Complex Problem Solving,
    LugerStubblefield (p.75-78).
  • Menyarankan investasi (saham/tabungan).
  • Jk. tabungan tidak cukup tingkatkan tabungan.
  • Jk. tabungan cukup pendapatan cukup saham
  • Jk. tabungan cukup pendapatan tdk. cukup
    dua-duanya.
  • Tabungan cukup gt 5jt. per dependent.
  • Penghslan cukup gt 12jt. per thn gt 4 jt./per
    dependent.
  • Buat representasi FOL.
  • Coba utk. suatu kasus.
Write a Comment
User Comments (0)
About PowerShow.com