Download presentation
Presentation is loading. Please wait.
Published byClarence Simpson Modified over 9 years ago
1
Knowledge Repn. & Reasoning Lec #11: Partitioning & Treewidth UIUC CS 498: Section EA Professor: Eyal Amir Fall Semester 2004
2
Last Time Resolution strategies –Deletion of clauses –Restricting resolution to some pairs –Ordering resolution between clauses Some strategies refutation complete, others only complete for Horn refutation
3
Today 1.We can partition reasoning while not hurting soundness and completeness 2.How to partition a KB with the best computational benefit Still maintaining soundness & completeness 3. Applications du jour: Planning
4
key locked can_open can_open open opened opened fetch broom key opened open opened broom fetch broom dry can_clean can_clean cleaned broom let_dry dry time let_dry drier let_dry time drier High-Level Structure in First- Order Logic
5
key locked can_open can_open open opened opened fetch broom key opened open opened broom fetch broom dry can_clean can_clean cleaned broom let_dry dry time let_dry drier let_dry time drier broom
6
Structured First-Order Reasoning Craig’s interpolation theorem (First-Order Logic): –If A B, then there is a formula C including only symbols from L(A) L(B) such that A C and C B clean
7
Structured First-Order Reasoning Craig’s interpolation theorem (First-Order Logic): –If A B, then there is a formula C including only symbols from L(A) L(B) such that A C and C B clean
8
High-Level Structure in First- Order Logic key locked can_open can_open open opened opened fetch broom key opened open opened broom fetch broom dry can_clean can_clean cleaned broom let_dry dry time let_dry drier let_dry time drier broom
9
Structured First-Order Reasoning Craig’s interpolation theorem (First-Order Logic): –If A B, then there is a formula C including only symbols from L(A) L(B) such that A C and C B clean broom
10
High-Level Structure in First- Order Logic key locked can_open can_open open opened opened fetch broom key opened open opened broom fetch broom dry can_clean can_clean cleaned broom let_dry dry time let_dry drier let_dry time drier broom
11
Start with a tree-decomposition partition graph Reasoning with partitions using MP MP Algorithm Pass messages in L i toward goal Identify goal partition Direct edges toward goal (fixing outbound link language L i for each partition) Concurrently, in each partition: Generate consequences in L i
12
Another Example Message Passing: Espresso machineEspresso machine SAT via partitioning
13
Benefits of Message-Passing Search space is restricted Allows parallel processing Sound and complete Can use different reasoners for each partition Small links imply short proofs Small partitions imply short proofs
14
High-Level Structure in First- Order Logic Has(key(x)) locked(x) can_open(x) can_open(x) open(x) opened(x) opened(x) fetch(y,x) in(y,x) has(y) Has(key(closet)) opened(closet) open(closet) opened(closet) In(broom,closet) fetch(broom,closet) Has(broom) dry(broom) can_clean(x) can_clean(x) cleaned(x) Has(y) let_dry(y) dry(y) Has(time) let_dry(y) Has(drier) let_dry(y) Has(time) Has(drier) Has broom
15
Structured First-Order Reasoning Craig’s interpolation theorem (First-Order Logic): –If A B, then there is a formula C including only symbols from L(A) L(B) such that A C and C B clean(room) Has(broom)
16
Structured First-Order Reasoning Craig’s interpolation theorem (First-Order Logic): –If A B, then there is a formula C including only symbols from L(A) L(B) such that A C and C B clean(room) Has(broom) x Has(x) Has(broom)
17
Contents 1.We can partition reasoning while not hurting soundness and completeness 2.How to partition a KB with the best computational benefit Still maintaining soundness & completeness 3. Applications: Planning
18
key locked can_open can_open open opened opened fetch broom key opened open opened broom fetch broom dry can_clean can_clean cleaned broom let_dry dry time let_dry drier let_dry time drier Automatic Decomposition of a Theory
19
key locked can_open can_open open opened opened fetch broom key opened open opened broom fetch broom dry can_clean can_clean cleaned broom let_dry dry time let_dry drier let_dry time drier Automatic Decomposition of a Theory
20
key locked can_open can_open open opened opened fetch broom key opened open opened broom fetch broom dry can_clean can_clean cleaned broom let_dry dry time let_dry drier let_dry time drier Automatic Decomposition of a Theory
21
keylocked can_openopen openedfetch broom dry cleaned let_dry time drier can_clean
22
Automatic Decomposition of a Theory keylocked can_openopen openedfetch broom dry cleaned let_dry time drier can_clean
23
Automatic Decomposition of a Theory keylocked can_openopen openedfetch broom dry cleaned let_dry time drier can_clean broom
24
Automatic Decomposition of a Theory broom keylocked can_openopen openedfetch broom dry cleaned let_dry time drier can_clean broom
25
Automatic Decomposition of a Theory key locked can_open can_open open opened opened fetch broom key opened open opened broom fetch broom dry can_clean can_clean cleaned broom let_dry dry time let_dry drier let_dry time drier broom
26
Automatic Partitioning Begin with a KB in PL or FOL Construct symbol graph –Edges join symbols which appear together in an axiom Find a tree decomposition of low width –Roughly, generalizes balanced vertex cut Partition axioms correspondingly –Each partition has its own vocabulary –Edge labels defined by shared vocabulary
27
Automatic Partitioning Find a tree decomposition of minimum width: –A tree in which each node corresponds to a set of vertices from the original graph –The tree satisfies the running intersection property: if v appears in two nodes in the tree, then v appears in all the nodes on the path connecting them –The width of the tree is the size of its largest node
28
Why Tree Decomposition? Example: BREAK-CYCLESBREAK-CYCLES
29
Automatic Partitioning Treewidth: [Robertson & Seymour ’86], … Approximation Algorithms: –General theories: [A. & McIlraith ’00] –O(Log(OPT))-approximation for general graphs: [A. ’01] –Constant factor approximation for planar graphs: [Seymour & Thomas ’94], [A., Krauthgamer & Rao ’03]
30
Automatic Partitioning: Heuristics Heuristic: min-degree 1.Given a graph G; List L - empty 2.Add to L a node v with minimum number of neighbors 3.Make a clique from v’s neighbors 4.Remove v from G 5.If G is empty, return L 6.Go to 2
31
Automatic Partitioning: Heuristics Heuristic: min-fill 1.Given a graph G; List L - empty 2.Add to L a node v with minimum number of edges missing between neighbors 3.Make a clique from v’s neighbors 4.Remove v from G 5.If G is empty, return L 6.Go to 2
32
Reasoning is performed locally in each partition Specialized reasoning procedures in every partition Globally sound & complete … provided each local reasoner is sound & complete for L i -consequence finding Performance is worst-case exponential within partitions, but linear in tree structure Summary: Characteristics of MP Minimizes between-partition deduction Supports parallel processing Different reasoners in different partitions Focuses within-partition deduction
33
Contents 1.We can partition reasoning while not hurting soundness and completeness 2.How to partition a KB with the best computational benefit Still maintaining soundness & completeness 3. Applications: Planning
34
Application: Planning General-purpose planning problem: –Given: Domain features (fluents) Action descriptions: effects, preconditions Initial state Goal condition –Find: Sequence of actions that is guaranteed to achieve the goal starting from the initial state
35
Application: Planning with partitions PartPlan Algorithm Start with a tree- structured partition graph Identify goal partition Direct edges toward goal In each partition –Generate all plans possible with depth d and width k –Pass messages toward goal
36
Planning with partitions PartPlan Algorithm Start with a tree-structured partition graph Identify goal partition Direct edges toward goal In each partition –Generate all plans possible with depth d and width k: “if you give me a block, I can return it to you painted”, “if you give me a block, let me do a few things, and then give me another block, then I can return the two painted and glued together.” –Pass messages toward goal: All preconditions/effects for which there are feasible action sequences
37
Factored Planning: Analysis Planner is sound and complete Running time for finding plans of width w with m partitions of treewidth k is O(m w 2 2w+2k ) Factoring can be done in polynomial time Goal can be distributed over partitions by adding at most 2 features per partition
38
Next Time Probabilistic Graphical Models: –Directed models: Bayesian Networks –Undirected models: Markov Fields Requires prior knowledge of: –Treewidth and graph algorithms –Probability theory
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.