SlideShare a Scribd company logo
1 of 21
Download to read offline
CMPUT 657
Heuristic Search
Final Project Presentation
Topic
A new Bi-directional A* search with Shorter Post
Processing on 8/15 puzzle
Presented By
Mohamad Saiful Islam
Outline
 Bi – Directional A* Search
 Balanced Algorithm
 New Bi-Directional A*
 Problem Domain
 Experimental Results
 Conclusion / Discussion
Bi-Directional A* Search
 Bi-Directional Search [1]
O Two simultaneous search processes starting from two
end points, root and goal
O Pohl [2] applied this to Dijkstra and A*
 Two Phases
O Main Phase: Two searches continue untill a meeting
point is found
O Post Phase: Make sure that path found is truely the
optimal one
 Post phase is needed as the first meeting point does
not gurantee optimal path
Bi-Directional A* Search
 Advantage
O A* space complexity two bi-dir search each
O Total much smaller than A*
 Pohl's [2] Conjecture – Bi-Directional A* can miss
the meeitng point, causing number of nodes
searched twice as big as A* [Missile Metaphor]
 Kaindl et. al [3] proved that the conjecture was
wrong, deviced an efficient implementation
Obd
 Obd/2

Balanced Heuristics
 Two search processes need two heuristics
O
 If A* reduces to Dijkstra
 Ikeda et. al[4] proved that A* becomes Dijkstra when
distance d is replaced by d' where
O
O So Bi-Directional A* becomes bi-directional Dijkstra
when
O
O Balanced Heuristics
O Bi-Directional A* using balanced heuristic considered to
be most effecient shortest path algorithms [4, 5, 6]
hu, hu
d'u,v=du,vhv−hu
huhu=constant
hu=0
New Bi-Directional Algorithm
 Node m scanned on both sides,
 For a node , the path from source to
destination through w is P

 Proved using the propety of Balanced heuristics
 So in post-phase, nodes that are only in the closed list
of opposite side are needed to be examined
 Bi-directional A* with balanced heuristics thus refrains
from examining huge number of nodes in post-
processing phase
L=gmgm
w∉S∪S
P ≥L
New Bi-directional Algorithm
 Let u is a node that will be added to the closed list and
the path through this node is P
O If then length of P is not smaller than L
O Then u can be rejected
O Proved using the propety of Balanced heuristics
 Algorithm using this in-equality can be relaxed from
using balanced heuristincs but they maintain the
short post phases of balanced heuristics
 New Algorithm proposed by Pijls at el[7] and
implemented in this project
guF−hu≥L
New Bi-directional Algorithm
1:forall v∈V do
2: gv=∞
3:endfor
4: S=∅
5: L=∞
6: gs=0 //sbecomesLABELED
7:boolean cand−found=true
8:while cand−found=true do
New Bi-directional Algorithm
9: C={v∣vislabelled∧gvhv−htL} //Cset ofall candidates
10: cand− found= false
11: while C≠∅∧cand− found=false do
12: uo=argmin{gvhv∣v∈C}
13: if uo∉S∧guoF−huo≥L then
14: C=C−{uo} // uo becomesREJECTED
15: else
16: cand−found=true //asuitablecandidateisfound
Trimming
Explicit Condition
Node not in opposite
closed list
17: endif
18: endwhile
19: if cand−found=true then
20: S=S{uo} // uo becomesSCANNED
21: F=guohuo
22: if uo∉S then
23: forall edgesuo,v∈E with vLABELED∨UNREACHED do
24: if gvguoduo ,v then
Nipping = Doubly scanned nodes
can not Expand more new nodes
25: gv=guoduo ,v //vbecomesrelabeled
26: predv=uo
27: L=min{L,gvgv}
28: endif
29: end for
30: end if
31: endif
32:endwhile
Problem Domain
 8/15 puzzle problem
O Class of sliding puzzle problem
 8 puzzle
O States, 181,440 unique states
 15 puzzle
O 10,461,394,944,000 unique states
 Most of the puzzle instances are hard to solve using A*
3.1×103
Experimental Result
 Heuristics
 Uni-directional
O Number of mis-placed tiles
O Manhattan Distance
O Manhattan Distance + Linear Conflict
 Third one performs the best [8] so used in the
Bi-Directional search as the base heuristic
Experimental Result
 Bi-directional
 Symmetric
O
 Balanced
O
hv=v ,t , hv=v ,t =s ,v
u ,v=under estimate of du ,v
hv=
1
2
{v ,t −s ,v}
hv=
1
2
{s ,v−v ,t}
Experimental Result
5 10 15 20 25 30 35
0
50000
100000
150000
200000
250000
300000
Number of States Generated
Misplaced Manhattan Balanced Symmetric
Solution Length
#States
Experimental Result
5 10 15 20 25 30 35
0
20000
40000
60000
80000
100000
120000
140000
160000
180000
Number of unique states generated
Misplaced
Manhattan
Balanced
Symmetric
Solution Length
#States
Experimental Result
Number States Number of Unique States Solution
Length
Misplac-
ed
Manha-
ttan
Balanc
ed
Symme-
tric
Misplaced Manhattan Balanced Symmetric
31189 5167 3557 855 27295 4814 3350 731 24
121498 12012 9323 2670 94075 11115 8676 2255 28
76238 3644 6555 518 63014 3452 6133 492 26
65565 7501 5615 2516 54493 6889 5258 2264 26
65750 9936 5514 1576 54794 9155 5160 1447 26
49548 6370 6024 1547 42021 5897 5632 1418 25
51165 11538 5957 1796 43087 10417 5574 1516 25
67277 8878 5663 1745 55841 8171 5305 1449 26
65715 5667 5400 1039 54647 5320 5066 896 26
49704 3680 5487 935 42520 3448 5139 872 25
Experimental Result
Manhattan Symmetric
# States # Unique States # States # Unique States Solution
Length
332474 313769 64963 52145 32
571221 526604 66103 52928 33
432482 407974 27726 22569 34
334703 310299 30438 26108 35
671954 625793 136149 110414 36
211219 199339 16620 12934 37
96297 80029 38
18102 15356 39
59120 48104 40
Conclusion
 Bi-Directional Search better than Uni-
Directional Search
 Symmetric Heuristics works better than
Balanced Heuristic
 Optimal Results
 Future work
O New Heuristics (landmark heuristic in Shortest
Path)
O New Domain
References
 [1] T.A.J. Nicholson, Finding the shortest route between two points in
a network, Computer Journal 9 (1966) 275–289.
 [2] I. Pohl, Bi-directional search, Machine Intelligence 6 (1971) 124–
140.
 [3] H. Kaindl, G. Kainz, Bidirectional heuristic search reconsidered, Journal
of Artificial Intelligence 38 (1) (1989) 95–109.
 [4] T.K. Ikeda, M. Hsu, H. Inai, S. Nishimura, H. Shimoura, T. Hashimoto,
K.Tenmoku, K. Mitoh, A fast algorithm for finding better routes by AI
search techniques, in: Proceedings Vehicle Navigation and Information
Systems Conference, IEEE, 1994.
 [5] A.V. Goldberg, C. Harrelson, Computing the shortest path: A* search
meets graph theory, in: 16th Annual ACM–SIAM Symposium on Discrete
Algorithms (SODA’05), 2005.
 [6] G.A. Klunder, H.N. Post, The shortest path problem on large scale real
road networks, Networks 48 (4) (2006) 182–194
 [7] W. Pijls, H. Post, A new bidirectional search algorithm with shortened
postprocessing, European Journal of Operational Research 198 (2009) 363–
369
References Cont..
 [8] O. Hansson, A. Mayer, and M. Yung, "Criticizing Solutions to Relaxed
Models Yields Powerful Admissible Heuristics," Information Sciences, Vol.
63, Issue 3, pp. 207-227, 1992.

More Related Content

What's hot

Uniformed tree searching
Uniformed tree searching Uniformed tree searching
Uniformed tree searching Ayaelshiwi
 
AI search techniques
AI search techniquesAI search techniques
AI search techniquesOmar Isaid
 
Agents in Artificial intelligence
Agents in Artificial intelligence Agents in Artificial intelligence
Agents in Artificial intelligence Lalit Birla
 
Gradient descent method
Gradient descent methodGradient descent method
Gradient descent methodSanghyuk Chun
 
Artificial Intelligence Notes Unit 1
Artificial Intelligence Notes Unit 1 Artificial Intelligence Notes Unit 1
Artificial Intelligence Notes Unit 1 DigiGurukul
 
Particle Swarm Optimization - PSO
Particle Swarm Optimization - PSOParticle Swarm Optimization - PSO
Particle Swarm Optimization - PSOMohamed Talaat
 
Depth First Search, Breadth First Search and Best First Search
Depth First Search, Breadth First Search and Best First SearchDepth First Search, Breadth First Search and Best First Search
Depth First Search, Breadth First Search and Best First SearchAdri Jovin
 
Uninformed search /Blind search in AI
Uninformed search /Blind search in AIUninformed search /Blind search in AI
Uninformed search /Blind search in AIKirti Verma
 
Lecture 23 alpha beta pruning
Lecture 23 alpha beta pruningLecture 23 alpha beta pruning
Lecture 23 alpha beta pruningHema Kashyap
 
Decision Tree Learning
Decision Tree LearningDecision Tree Learning
Decision Tree LearningMilind Gokhale
 
Hill climbing algorithm in artificial intelligence
Hill climbing algorithm in artificial intelligenceHill climbing algorithm in artificial intelligence
Hill climbing algorithm in artificial intelligencesandeep54552
 
Artificial Intelligence Searching Techniques
Artificial Intelligence Searching TechniquesArtificial Intelligence Searching Techniques
Artificial Intelligence Searching TechniquesDr. C.V. Suresh Babu
 
Artificial Intelligence Game Search by Examples
Artificial Intelligence Game Search by ExamplesArtificial Intelligence Game Search by Examples
Artificial Intelligence Game Search by ExamplesAhmed Gad
 
Inference in Bayesian Networks
Inference in Bayesian NetworksInference in Bayesian Networks
Inference in Bayesian Networksguestfee8698
 
Evolutionary Algorithms
Evolutionary AlgorithmsEvolutionary Algorithms
Evolutionary AlgorithmsReem Alattas
 
Lecture 14 Heuristic Search-A star algorithm
Lecture 14 Heuristic Search-A star algorithmLecture 14 Heuristic Search-A star algorithm
Lecture 14 Heuristic Search-A star algorithmHema Kashyap
 
Uninformed Search technique
Uninformed Search techniqueUninformed Search technique
Uninformed Search techniqueKapil Dahal
 

What's hot (20)

Uniformed tree searching
Uniformed tree searching Uniformed tree searching
Uniformed tree searching
 
AI search techniques
AI search techniquesAI search techniques
AI search techniques
 
Agents in Artificial intelligence
Agents in Artificial intelligence Agents in Artificial intelligence
Agents in Artificial intelligence
 
Gradient descent method
Gradient descent methodGradient descent method
Gradient descent method
 
Daa notes 3
Daa notes 3Daa notes 3
Daa notes 3
 
Artificial Intelligence Notes Unit 1
Artificial Intelligence Notes Unit 1 Artificial Intelligence Notes Unit 1
Artificial Intelligence Notes Unit 1
 
Particle Swarm Optimization - PSO
Particle Swarm Optimization - PSOParticle Swarm Optimization - PSO
Particle Swarm Optimization - PSO
 
Depth First Search, Breadth First Search and Best First Search
Depth First Search, Breadth First Search and Best First SearchDepth First Search, Breadth First Search and Best First Search
Depth First Search, Breadth First Search and Best First Search
 
Uninformed search /Blind search in AI
Uninformed search /Blind search in AIUninformed search /Blind search in AI
Uninformed search /Blind search in AI
 
Lecture 23 alpha beta pruning
Lecture 23 alpha beta pruningLecture 23 alpha beta pruning
Lecture 23 alpha beta pruning
 
Decision Tree Learning
Decision Tree LearningDecision Tree Learning
Decision Tree Learning
 
Hill climbing algorithm in artificial intelligence
Hill climbing algorithm in artificial intelligenceHill climbing algorithm in artificial intelligence
Hill climbing algorithm in artificial intelligence
 
Artificial Intelligence Searching Techniques
Artificial Intelligence Searching TechniquesArtificial Intelligence Searching Techniques
Artificial Intelligence Searching Techniques
 
Artificial Intelligence Game Search by Examples
Artificial Intelligence Game Search by ExamplesArtificial Intelligence Game Search by Examples
Artificial Intelligence Game Search by Examples
 
Inference in Bayesian Networks
Inference in Bayesian NetworksInference in Bayesian Networks
Inference in Bayesian Networks
 
Evolutionary Algorithms
Evolutionary AlgorithmsEvolutionary Algorithms
Evolutionary Algorithms
 
A* Search Algorithm
A* Search AlgorithmA* Search Algorithm
A* Search Algorithm
 
K Nearest Neighbors
K Nearest NeighborsK Nearest Neighbors
K Nearest Neighbors
 
Lecture 14 Heuristic Search-A star algorithm
Lecture 14 Heuristic Search-A star algorithmLecture 14 Heuristic Search-A star algorithm
Lecture 14 Heuristic Search-A star algorithm
 
Uninformed Search technique
Uninformed Search techniqueUninformed Search technique
Uninformed Search technique
 

Viewers also liked

WellData and WellData Real-Time systems Flyer
WellData and WellData Real-Time systems FlyerWellData and WellData Real-Time systems Flyer
WellData and WellData Real-Time systems FlyerMohammad Saiful Islam
 
Cocoaheads - Streaming on iOS devices
Cocoaheads - Streaming on iOS devicesCocoaheads - Streaming on iOS devices
Cocoaheads - Streaming on iOS devicesNirbhay Kundan
 
Require JS
Require JSRequire JS
Require JSImaginea
 
Travel Plan using Geo-tagged Photos in Geocrowd2013
Travel Plan using Geo-tagged Photos in Geocrowd2013 Travel Plan using Geo-tagged Photos in Geocrowd2013
Travel Plan using Geo-tagged Photos in Geocrowd2013 Arizona State University
 
Technology Seminar Handout
Technology Seminar HandoutTechnology Seminar Handout
Technology Seminar HandoutDerecskei Anita
 
Seminar "Technology Enhanced Learning" - Einheit 1
Seminar "Technology Enhanced Learning" - Einheit 1Seminar "Technology Enhanced Learning" - Einheit 1
Seminar "Technology Enhanced Learning" - Einheit 1Martin Ebner
 
Heuristics
HeuristicsHeuristics
HeuristicsMoe Tmz
 
Heuristc Search Techniques
Heuristc Search TechniquesHeuristc Search Techniques
Heuristc Search TechniquesJismy .K.Jose
 
16890 unit 2 heuristic search techniques
16890 unit 2 heuristic  search techniques16890 unit 2 heuristic  search techniques
16890 unit 2 heuristic search techniquesJais Balta
 
Lecture 11 Informed Search
Lecture 11 Informed SearchLecture 11 Informed Search
Lecture 11 Informed SearchHema Kashyap
 
Solving problems by searching Informed (heuristics) Search
Solving problems by searching Informed (heuristics) SearchSolving problems by searching Informed (heuristics) Search
Solving problems by searching Informed (heuristics) Searchmatele41
 
Optimization Heuristics
Optimization HeuristicsOptimization Heuristics
Optimization HeuristicsKausal Malladi
 
Firefly algorithm
Firefly algorithmFirefly algorithm
Firefly algorithmHasan Gök
 
09 heuristic search
09 heuristic search09 heuristic search
09 heuristic searchTianlu Wang
 
(Radhika) presentation on chapter 2 ai
(Radhika) presentation on chapter 2 ai(Radhika) presentation on chapter 2 ai
(Radhika) presentation on chapter 2 aiRadhika Srinivasan
 

Viewers also liked (20)

WellData and WellData Real-Time systems Flyer
WellData and WellData Real-Time systems FlyerWellData and WellData Real-Time systems Flyer
WellData and WellData Real-Time systems Flyer
 
Cocoaheads - Streaming on iOS devices
Cocoaheads - Streaming on iOS devicesCocoaheads - Streaming on iOS devices
Cocoaheads - Streaming on iOS devices
 
Require js
Require jsRequire js
Require js
 
Require JS
Require JSRequire JS
Require JS
 
Travel Plan using Geo-tagged Photos in Geocrowd2013
Travel Plan using Geo-tagged Photos in Geocrowd2013 Travel Plan using Geo-tagged Photos in Geocrowd2013
Travel Plan using Geo-tagged Photos in Geocrowd2013
 
Technology Seminar Handout
Technology Seminar HandoutTechnology Seminar Handout
Technology Seminar Handout
 
simple
simplesimple
simple
 
Seminar "Technology Enhanced Learning" - Einheit 1
Seminar "Technology Enhanced Learning" - Einheit 1Seminar "Technology Enhanced Learning" - Einheit 1
Seminar "Technology Enhanced Learning" - Einheit 1
 
Heuristics
HeuristicsHeuristics
Heuristics
 
M4 Heuristics
M4 HeuristicsM4 Heuristics
M4 Heuristics
 
Heuristc Search Techniques
Heuristc Search TechniquesHeuristc Search Techniques
Heuristc Search Techniques
 
16890 unit 2 heuristic search techniques
16890 unit 2 heuristic  search techniques16890 unit 2 heuristic  search techniques
16890 unit 2 heuristic search techniques
 
04 search heuristic
04 search heuristic04 search heuristic
04 search heuristic
 
Lecture 11 Informed Search
Lecture 11 Informed SearchLecture 11 Informed Search
Lecture 11 Informed Search
 
Solving problems by searching Informed (heuristics) Search
Solving problems by searching Informed (heuristics) SearchSolving problems by searching Informed (heuristics) Search
Solving problems by searching Informed (heuristics) Search
 
Optimization Heuristics
Optimization HeuristicsOptimization Heuristics
Optimization Heuristics
 
Lect6 csp
Lect6 cspLect6 csp
Lect6 csp
 
Firefly algorithm
Firefly algorithmFirefly algorithm
Firefly algorithm
 
09 heuristic search
09 heuristic search09 heuristic search
09 heuristic search
 
(Radhika) presentation on chapter 2 ai
(Radhika) presentation on chapter 2 ai(Radhika) presentation on chapter 2 ai
(Radhika) presentation on chapter 2 ai
 

Similar to Presentation - Bi-directional A-star search

AI3391 ARTIFICIAL INTELLIGENCE UNIT II notes.pdf
AI3391 ARTIFICIAL INTELLIGENCE UNIT II notes.pdfAI3391 ARTIFICIAL INTELLIGENCE UNIT II notes.pdf
AI3391 ARTIFICIAL INTELLIGENCE UNIT II notes.pdfAsst.prof M.Gokilavani
 
AI_Session 8 A searching algorithm .pptx
AI_Session 8 A searching algorithm .pptxAI_Session 8 A searching algorithm .pptx
AI_Session 8 A searching algorithm .pptxAsst.prof M.Gokilavani
 
Seminar Report (Final)
Seminar Report (Final)Seminar Report (Final)
Seminar Report (Final)Aruneel Das
 
1984 Article on An Application of AI to Operations Reserach
1984 Article on An Application of AI to Operations Reserach1984 Article on An Application of AI to Operations Reserach
1984 Article on An Application of AI to Operations ReserachBob Marcus
 
AI3391 Session 10 A searching algorithm.pptx
AI3391 Session 10 A searching algorithm.pptxAI3391 Session 10 A searching algorithm.pptx
AI3391 Session 10 A searching algorithm.pptxAsst.prof M.Gokilavani
 
Introduction to computing Processing and performance.pdf
Introduction to computing Processing and performance.pdfIntroduction to computing Processing and performance.pdf
Introduction to computing Processing and performance.pdfTulasiramKandula1
 
DSA Complexity.pptx What is Complexity Analysis? What is the need for Compl...
DSA Complexity.pptx   What is Complexity Analysis? What is the need for Compl...DSA Complexity.pptx   What is Complexity Analysis? What is the need for Compl...
DSA Complexity.pptx What is Complexity Analysis? What is the need for Compl...2022cspaawan12556
 
Point Placement Algorithms: An Experimental Study
Point Placement Algorithms: An Experimental StudyPoint Placement Algorithms: An Experimental Study
Point Placement Algorithms: An Experimental StudyCSCJournals
 
Branch and bounding : Data structures
Branch and bounding : Data structuresBranch and bounding : Data structures
Branch and bounding : Data structuresKàŕtheek Jåvvàjí
 
Heuristic Function Influence to the Global Optimum Value in Shortest Path Pro...
Heuristic Function Influence to the Global Optimum Value in Shortest Path Pro...Heuristic Function Influence to the Global Optimum Value in Shortest Path Pro...
Heuristic Function Influence to the Global Optimum Value in Shortest Path Pro...Universitas Pembangunan Panca Budi
 
Single_Variable_Optimization_Part1_Dichotomous_moodle.ppsx
Single_Variable_Optimization_Part1_Dichotomous_moodle.ppsxSingle_Variable_Optimization_Part1_Dichotomous_moodle.ppsx
Single_Variable_Optimization_Part1_Dichotomous_moodle.ppsxAmerTout
 
Discrete Math IP4 - Automata Theory
Discrete Math IP4 - Automata TheoryDiscrete Math IP4 - Automata Theory
Discrete Math IP4 - Automata TheoryMark Simon
 
Bidirectional graph search techniques for finding shortest path in image base...
Bidirectional graph search techniques for finding shortest path in image base...Bidirectional graph search techniques for finding shortest path in image base...
Bidirectional graph search techniques for finding shortest path in image base...Navin Kumar
 
Ip 5 discrete mathematics
Ip 5 discrete mathematicsIp 5 discrete mathematics
Ip 5 discrete mathematicsMark Simon
 
Global Optimization with Descending Region Algorithm
Global Optimization with Descending Region AlgorithmGlobal Optimization with Descending Region Algorithm
Global Optimization with Descending Region AlgorithmLoc Nguyen
 
Selection K in K-means Clustering
Selection K in K-means ClusteringSelection K in K-means Clustering
Selection K in K-means ClusteringJunghoon Kim
 
Design and Analysis of Algorithms Lecture Notes
Design and Analysis of Algorithms Lecture NotesDesign and Analysis of Algorithms Lecture Notes
Design and Analysis of Algorithms Lecture NotesSreedhar Chowdam
 

Similar to Presentation - Bi-directional A-star search (20)

AI3391 ARTIFICIAL INTELLIGENCE UNIT II notes.pdf
AI3391 ARTIFICIAL INTELLIGENCE UNIT II notes.pdfAI3391 ARTIFICIAL INTELLIGENCE UNIT II notes.pdf
AI3391 ARTIFICIAL INTELLIGENCE UNIT II notes.pdf
 
AI_Session 8 A searching algorithm .pptx
AI_Session 8 A searching algorithm .pptxAI_Session 8 A searching algorithm .pptx
AI_Session 8 A searching algorithm .pptx
 
Seminar Report (Final)
Seminar Report (Final)Seminar Report (Final)
Seminar Report (Final)
 
1984 Article on An Application of AI to Operations Reserach
1984 Article on An Application of AI to Operations Reserach1984 Article on An Application of AI to Operations Reserach
1984 Article on An Application of AI to Operations Reserach
 
AI3391 Session 10 A searching algorithm.pptx
AI3391 Session 10 A searching algorithm.pptxAI3391 Session 10 A searching algorithm.pptx
AI3391 Session 10 A searching algorithm.pptx
 
Introduction to computing Processing and performance.pdf
Introduction to computing Processing and performance.pdfIntroduction to computing Processing and performance.pdf
Introduction to computing Processing and performance.pdf
 
DSA Complexity.pptx What is Complexity Analysis? What is the need for Compl...
DSA Complexity.pptx   What is Complexity Analysis? What is the need for Compl...DSA Complexity.pptx   What is Complexity Analysis? What is the need for Compl...
DSA Complexity.pptx What is Complexity Analysis? What is the need for Compl...
 
Point Placement Algorithms: An Experimental Study
Point Placement Algorithms: An Experimental StudyPoint Placement Algorithms: An Experimental Study
Point Placement Algorithms: An Experimental Study
 
Branch and bounding : Data structures
Branch and bounding : Data structuresBranch and bounding : Data structures
Branch and bounding : Data structures
 
Heuristic Function Influence to the Global Optimum Value in Shortest Path Pro...
Heuristic Function Influence to the Global Optimum Value in Shortest Path Pro...Heuristic Function Influence to the Global Optimum Value in Shortest Path Pro...
Heuristic Function Influence to the Global Optimum Value in Shortest Path Pro...
 
A greedy algorithms
A greedy algorithmsA greedy algorithms
A greedy algorithms
 
Single_Variable_Optimization_Part1_Dichotomous_moodle.ppsx
Single_Variable_Optimization_Part1_Dichotomous_moodle.ppsxSingle_Variable_Optimization_Part1_Dichotomous_moodle.ppsx
Single_Variable_Optimization_Part1_Dichotomous_moodle.ppsx
 
Presentation sim opencv
Presentation sim opencvPresentation sim opencv
Presentation sim opencv
 
Discrete Math IP4 - Automata Theory
Discrete Math IP4 - Automata TheoryDiscrete Math IP4 - Automata Theory
Discrete Math IP4 - Automata Theory
 
Bidirectional graph search techniques for finding shortest path in image base...
Bidirectional graph search techniques for finding shortest path in image base...Bidirectional graph search techniques for finding shortest path in image base...
Bidirectional graph search techniques for finding shortest path in image base...
 
Ip 5 discrete mathematics
Ip 5 discrete mathematicsIp 5 discrete mathematics
Ip 5 discrete mathematics
 
Global Optimization with Descending Region Algorithm
Global Optimization with Descending Region AlgorithmGlobal Optimization with Descending Region Algorithm
Global Optimization with Descending Region Algorithm
 
Selection K in K-means Clustering
Selection K in K-means ClusteringSelection K in K-means Clustering
Selection K in K-means Clustering
 
Design and Analysis of Algorithms Lecture Notes
Design and Analysis of Algorithms Lecture NotesDesign and Analysis of Algorithms Lecture Notes
Design and Analysis of Algorithms Lecture Notes
 
36 greedy
36 greedy36 greedy
36 greedy
 

Presentation - Bi-directional A-star search

  • 1. CMPUT 657 Heuristic Search Final Project Presentation Topic A new Bi-directional A* search with Shorter Post Processing on 8/15 puzzle Presented By Mohamad Saiful Islam
  • 2. Outline  Bi – Directional A* Search  Balanced Algorithm  New Bi-Directional A*  Problem Domain  Experimental Results  Conclusion / Discussion
  • 3. Bi-Directional A* Search  Bi-Directional Search [1] O Two simultaneous search processes starting from two end points, root and goal O Pohl [2] applied this to Dijkstra and A*  Two Phases O Main Phase: Two searches continue untill a meeting point is found O Post Phase: Make sure that path found is truely the optimal one  Post phase is needed as the first meeting point does not gurantee optimal path
  • 4. Bi-Directional A* Search  Advantage O A* space complexity two bi-dir search each O Total much smaller than A*  Pohl's [2] Conjecture – Bi-Directional A* can miss the meeitng point, causing number of nodes searched twice as big as A* [Missile Metaphor]  Kaindl et. al [3] proved that the conjecture was wrong, deviced an efficient implementation Obd  Obd/2 
  • 5. Balanced Heuristics  Two search processes need two heuristics O  If A* reduces to Dijkstra  Ikeda et. al[4] proved that A* becomes Dijkstra when distance d is replaced by d' where O O So Bi-Directional A* becomes bi-directional Dijkstra when O O Balanced Heuristics O Bi-Directional A* using balanced heuristic considered to be most effecient shortest path algorithms [4, 5, 6] hu, hu d'u,v=du,vhv−hu huhu=constant hu=0
  • 6. New Bi-Directional Algorithm  Node m scanned on both sides,  For a node , the path from source to destination through w is P   Proved using the propety of Balanced heuristics  So in post-phase, nodes that are only in the closed list of opposite side are needed to be examined  Bi-directional A* with balanced heuristics thus refrains from examining huge number of nodes in post- processing phase L=gmgm w∉S∪S P ≥L
  • 7. New Bi-directional Algorithm  Let u is a node that will be added to the closed list and the path through this node is P O If then length of P is not smaller than L O Then u can be rejected O Proved using the propety of Balanced heuristics  Algorithm using this in-equality can be relaxed from using balanced heuristincs but they maintain the short post phases of balanced heuristics  New Algorithm proposed by Pijls at el[7] and implemented in this project guF−hu≥L
  • 8. New Bi-directional Algorithm 1:forall v∈V do 2: gv=∞ 3:endfor 4: S=∅ 5: L=∞ 6: gs=0 //sbecomesLABELED 7:boolean cand−found=true 8:while cand−found=true do
  • 9. New Bi-directional Algorithm 9: C={v∣vislabelled∧gvhv−htL} //Cset ofall candidates 10: cand− found= false 11: while C≠∅∧cand− found=false do 12: uo=argmin{gvhv∣v∈C} 13: if uo∉S∧guoF−huo≥L then 14: C=C−{uo} // uo becomesREJECTED 15: else 16: cand−found=true //asuitablecandidateisfound Trimming Explicit Condition Node not in opposite closed list
  • 10. 17: endif 18: endwhile 19: if cand−found=true then 20: S=S{uo} // uo becomesSCANNED 21: F=guohuo 22: if uo∉S then 23: forall edgesuo,v∈E with vLABELED∨UNREACHED do 24: if gvguoduo ,v then Nipping = Doubly scanned nodes can not Expand more new nodes
  • 11. 25: gv=guoduo ,v //vbecomesrelabeled 26: predv=uo 27: L=min{L,gvgv} 28: endif 29: end for 30: end if 31: endif 32:endwhile
  • 12. Problem Domain  8/15 puzzle problem O Class of sliding puzzle problem  8 puzzle O States, 181,440 unique states  15 puzzle O 10,461,394,944,000 unique states  Most of the puzzle instances are hard to solve using A* 3.1×103
  • 13. Experimental Result  Heuristics  Uni-directional O Number of mis-placed tiles O Manhattan Distance O Manhattan Distance + Linear Conflict  Third one performs the best [8] so used in the Bi-Directional search as the base heuristic
  • 14. Experimental Result  Bi-directional  Symmetric O  Balanced O hv=v ,t , hv=v ,t =s ,v u ,v=under estimate of du ,v hv= 1 2 {v ,t −s ,v} hv= 1 2 {s ,v−v ,t}
  • 15. Experimental Result 5 10 15 20 25 30 35 0 50000 100000 150000 200000 250000 300000 Number of States Generated Misplaced Manhattan Balanced Symmetric Solution Length #States
  • 16. Experimental Result 5 10 15 20 25 30 35 0 20000 40000 60000 80000 100000 120000 140000 160000 180000 Number of unique states generated Misplaced Manhattan Balanced Symmetric Solution Length #States
  • 17. Experimental Result Number States Number of Unique States Solution Length Misplac- ed Manha- ttan Balanc ed Symme- tric Misplaced Manhattan Balanced Symmetric 31189 5167 3557 855 27295 4814 3350 731 24 121498 12012 9323 2670 94075 11115 8676 2255 28 76238 3644 6555 518 63014 3452 6133 492 26 65565 7501 5615 2516 54493 6889 5258 2264 26 65750 9936 5514 1576 54794 9155 5160 1447 26 49548 6370 6024 1547 42021 5897 5632 1418 25 51165 11538 5957 1796 43087 10417 5574 1516 25 67277 8878 5663 1745 55841 8171 5305 1449 26 65715 5667 5400 1039 54647 5320 5066 896 26 49704 3680 5487 935 42520 3448 5139 872 25
  • 18. Experimental Result Manhattan Symmetric # States # Unique States # States # Unique States Solution Length 332474 313769 64963 52145 32 571221 526604 66103 52928 33 432482 407974 27726 22569 34 334703 310299 30438 26108 35 671954 625793 136149 110414 36 211219 199339 16620 12934 37 96297 80029 38 18102 15356 39 59120 48104 40
  • 19. Conclusion  Bi-Directional Search better than Uni- Directional Search  Symmetric Heuristics works better than Balanced Heuristic  Optimal Results  Future work O New Heuristics (landmark heuristic in Shortest Path) O New Domain
  • 20. References  [1] T.A.J. Nicholson, Finding the shortest route between two points in a network, Computer Journal 9 (1966) 275–289.  [2] I. Pohl, Bi-directional search, Machine Intelligence 6 (1971) 124– 140.  [3] H. Kaindl, G. Kainz, Bidirectional heuristic search reconsidered, Journal of Artificial Intelligence 38 (1) (1989) 95–109.  [4] T.K. Ikeda, M. Hsu, H. Inai, S. Nishimura, H. Shimoura, T. Hashimoto, K.Tenmoku, K. Mitoh, A fast algorithm for finding better routes by AI search techniques, in: Proceedings Vehicle Navigation and Information Systems Conference, IEEE, 1994.  [5] A.V. Goldberg, C. Harrelson, Computing the shortest path: A* search meets graph theory, in: 16th Annual ACM–SIAM Symposium on Discrete Algorithms (SODA’05), 2005.  [6] G.A. Klunder, H.N. Post, The shortest path problem on large scale real road networks, Networks 48 (4) (2006) 182–194  [7] W. Pijls, H. Post, A new bidirectional search algorithm with shortened postprocessing, European Journal of Operational Research 198 (2009) 363– 369
  • 21. References Cont..  [8] O. Hansson, A. Mayer, and M. Yung, "Criticizing Solutions to Relaxed Models Yields Powerful Admissible Heuristics," Information Sciences, Vol. 63, Issue 3, pp. 207-227, 1992.