04_LocalSearch - Local Search Scaling Up So far, we have...

Info iconThis preview shows pages 1–19. Sign up to view the full content.

View Full Document Right Arrow Icon
Local Search
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Scaling Up So far, we have considered methods that systematically explore the full search space, possibly using principled pruning (A* etc.). The current best such algorithms (RBFS / SMA*) can handle search spaces of up to 10 100 states → ~ 500 binary valued variables. But search spaces for some real-world problems might be much bigger - e.g. 10 30,000 states. Here, a completely different kind of search is needed. Local Search Methods
Background image of page 2
Optimization Problems We're interested in the Goal State - not in how to get there. Optimization Problem: - State: vector of variables - Objective Function: f : state - Goal: find state that maximizes or minimizes the objective function Examples: VLSI layout, job scheduling, map coloring, N-Queens
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Representations for 8Q problem
Background image of page 4
Heuristic for 8Q problem?
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Heuristic for 8Q problem?
Background image of page 6
Local Search Methods Applicable to optimization problems. Basic idea: - use a single current state - don't save paths followed - generally move only to successors/neighbors of that state Generally require a complete state description .
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Hill-Climbing Search function HILL-CLIMBING ( ) returns a solution state inputs: , a problem static: , a node MAKE-NODE(INITIAL-STATE[ ]) loop do a highest-valued succe problem problem current current problem next ssor of if VALUE[next] < VALUE[current] then return end current current current next
Background image of page 8
Current State Evaluation
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Hill Climbing Pathologies Objective function Shoulder Global Maximum Local Maximum “flat” local maximum State Space Value of current solution
Background image of page 10
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 12
Local Maximum Example
Background image of page 13

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Neutral “Sideways” moves Take new state even if not strictly better (just equal) Allows exploring plateaus …But can get into cycles
Background image of page 14
Neutral “Sideways” moves 8Q problem without sideways Stuck 86% time 4 steps to succeed 3 to get stuck 8Q problem with sideways Succeeds 94% time 21 steps to succeed 64 to get stuck
Background image of page 15

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Random restarts Random restarts: Simply restart at a new random state after a pre-defined number of steps. Is it worth it? If probability of success is p, then Expected number of trials to success is 1/p
Background image of page 16
Neutral “Sideways” moves 8Q problem without sideways Stuck 86% time 4 steps to succeed 3 to get stuck 8Q problem with sideways Succeeds 94% time 21 steps to succeed 64 to get stuck A = With Sideways B = Without Sideways
Background image of page 17

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Nelder-Mead (Simplex) Method Reflect the point with the highest WSS through centroid (center) of the simplex If this produces the lowest WSS (best point) expand the simplex and reflect further If this is just a good point start at the top and reflect again
Background image of page 18
Image of page 19
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 45

04_LocalSearch - Local Search Scaling Up So far, we have...

This preview shows document pages 1 - 19. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online