Scribd
Upload
14 views4 pages

AI Lab 6 Manual

The document provides a manual for Lab 06 in an Artificial Intelligence course, focusing on two optimization algorithms: Simulated Annealing and Local Beam Search. Simulated Annealing is a stochastic method inspired by metallurgy that allows for occasional uphill moves to escape local optima, while Local Beam Search explores multiple candidate solutions simultaneously to enhance the search process. Each algorithm is detailed with step-by-step workings and components necessary for implementation.

Uploaded by

i220818
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
14 views4 pages

AI Lab 6 Manual

The document provides a manual for Lab 06 in an Artificial Intelligence course, focusing on two optimization algorithms: Simulated Annealing and Local Beam Search. Simulated Annealing is a stochastic method inspired by metallurgy that allows for occasional uphill moves to escape local optima, while Local Beam Search explores multiple candidate solutions simultaneously to enhance the search process. Each algorithm is detailed with step-by-step workings and components necessary for implementation.

Uploaded by

i220818
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Artificial Intelligence

BS (CS) _SP_2025

Lab_06 Manual

Learning Objectives:

1. Simulated Annealing
2. Local Beam Search
1. Simulated Annealing:
Simulated annealing is a stochastic optimization algorithm that is often used to find the global
minimum (or maximum) of a non-convex function. The algorithm is inspired by the process of
annealing in metallurgy, where a metal is heated and then slowly cooled to reduce its defects and
improve its properties.

Simulated Annealing explores the solution space by allowing occasional uphill moves (worse
solutions) to escape local optima. This is controlled by a temperature parameter that decreases
over time, reducing the likelihood of accepting worse solutions as the search progresses.

Step-by-Step Working:
1. Initialize: Start with an initial solution and a high temperature.
2. Evaluate: Compute the cost of current solution.
3. Generate neighbor: Select a neighboring solution randomly.
4. Compute cost difference: measure the difference between new solution and current
solution.
5. Acceptance Criteria:
• if new solution is better, accept it.
• if new solution is worse, accept it with a probability dependent on
temperature.
6. Cool Down: reduce the temperature gradually.
7. Repeat: continue until the system is sufficiently cooled down or a stopping
condition is met.

Components of Simulated Annealing:


• Initial Solution → a random point in search space.
• Neighbor Selection → a method to select next solution/node.
• Cost Function → method to evaluate the quality of solution.
• Acceptance probability → decides whether to move to a worse solution or not using
Boltzmann probability:
• Cooling Schedule → determines how temperature decreases over time. Usually changed
slowly by a fixed fraction.
Simulated_Annealing (Problem, Initial_Temperature, Cooling_Rate)
current_solution = Initial_Solution(Problem) --- randomly generated initial solution
current_temperature = Initial_Temperature
while stopping condition not met do
new_solution = Generate_Neighbor(current_solution)
cost_difference = Cost(new_solution) - Cost(current_solution)
if cost_difference < 0 then
current_solution = new_solution
else
probability = exp (-cost_difference / current_temperature)
if Random (0,1) < probability then
current_solution = new_solution
current_temperature = current_temperature * Cooling_Rate
return current_solution

2. Local Beam Search:


Local Beam Search is a heuristic search algorithm that explores multiple candidate solutions
simultaneously. Unlike Simulated Annealing or Hill Climbing, which follow a single solution
path, Local Beam Search maintains a fixed number of "k" best solutions and expands them in
parallel. This approach helps in escaping local optima by leveraging information from multiple
promising solutions.

Step-by-Step Working:

1. Initialize: Start with "k" randomly generated solutions.


2. Evaluate: Compute the cost (or heuristic value) for each solution.
3. Generate Successors: Expand each solution to produce new neighboring states.
4. Select Best k Candidates: From all generated successors, keep only the "k" best
solutions.
5. Repeat: Continue until a stopping condition is met (e.g., a solution is good enough or a
maximum number of iterations is reached).
Local_Beam_Search(Problem, k)

Initialize k random solutions

while stopping condition not met do

Generate all possible successors of the k solutions

Select the best k successors based on evaluation function

return best found solution

You might also like