Artificial Intelligence, a modern approach
Chapter 4 with questions and well verified
answers actual exam!!! 2026
Local search and optimization problems - ANSWER -•Hill-climbing search
•Simulated annealing search
•Local beam search
•Genetic algorithms
In local search problems, state space = set of "complete" configurations; the goal is to find best
configuration satisfying constraints.
•These problems are unlike the search problems from previous classes:
- The path to the goal is irrelevant
-- all you care about is the final configuration
- These are often optimization problems in which you find the best state according to an
objective function
local search algorithms - ANSWER -•Two strategies for choosing the state to visit next
•Hill climbing
•Simulated annealing
•Then, an extension to multiple current states
:•Local Beam search
•Genetic algorithms
, Hill Climbing (Greedy Local Search) - ANSWER -•Start wherever
•Pick the best and replace the current state with that one
.•Loop
•If no neighbors better than current, quit
Don't look ahead beyond the immediate neighbors of the current state.
Objective function - ANSWER -Consider the states of a problem laid out in a state-space
landscape
To find the highest peak-a global maximum, the process is called hill climbing
To find the lowest value (e.g., cost)- global minimum, the process is gradient descent
Problems with Hill climbing search - ANSWER -Local Maximum: a peak that is lower than
the highest peak, so a suboptimal solution is returned
Plateau: the evaluation function is flat, resulting in random walk
Ridges: every neighbor seems to be a downhill, but search space has an uphill(just not in
neighbors)
Variants of hill climbing search - ANSWER -(designed to solve local optima problem)
stochastic hill climbing
-randomly select among better neighbors
converges more slowly, but finds better solutions in some landscapes
BUT WHAT IF THE NEIGHBORHOOD IS TOO LARGE TO ENUMERATE?
First choice hill climbing:
Chapter 4 with questions and well verified
answers actual exam!!! 2026
Local search and optimization problems - ANSWER -•Hill-climbing search
•Simulated annealing search
•Local beam search
•Genetic algorithms
In local search problems, state space = set of "complete" configurations; the goal is to find best
configuration satisfying constraints.
•These problems are unlike the search problems from previous classes:
- The path to the goal is irrelevant
-- all you care about is the final configuration
- These are often optimization problems in which you find the best state according to an
objective function
local search algorithms - ANSWER -•Two strategies for choosing the state to visit next
•Hill climbing
•Simulated annealing
•Then, an extension to multiple current states
:•Local Beam search
•Genetic algorithms
, Hill Climbing (Greedy Local Search) - ANSWER -•Start wherever
•Pick the best and replace the current state with that one
.•Loop
•If no neighbors better than current, quit
Don't look ahead beyond the immediate neighbors of the current state.
Objective function - ANSWER -Consider the states of a problem laid out in a state-space
landscape
To find the highest peak-a global maximum, the process is called hill climbing
To find the lowest value (e.g., cost)- global minimum, the process is gradient descent
Problems with Hill climbing search - ANSWER -Local Maximum: a peak that is lower than
the highest peak, so a suboptimal solution is returned
Plateau: the evaluation function is flat, resulting in random walk
Ridges: every neighbor seems to be a downhill, but search space has an uphill(just not in
neighbors)
Variants of hill climbing search - ANSWER -(designed to solve local optima problem)
stochastic hill climbing
-randomly select among better neighbors
converges more slowly, but finds better solutions in some landscapes
BUT WHAT IF THE NEIGHBORHOOD IS TOO LARGE TO ENUMERATE?
First choice hill climbing:
