T/F & Short answer

A finite state space always leads to a finite search tree. (T/F)

F

A* search can guarantee to find the optimal solution no matter if the state space has identical or varying step cost. (T/F)

F

An agent that senses only partial info about the state cannot be perfectly rational. (T/F)

F

Depth-first search always expands at least as many nodes as A* search with an admissible heuristic. (T/F)

F

Uniform-cost search is optimal even with non-identical step costs that are greater than 0.

F

Consider the environment for playing chess with a clock, the env is fully observable and deterministic. (T/F)

F clock

If the env is deterministic except for the for the actions of other agents, then its semidynamic. (T/F)

F its strategic

Tree search algorithm using iterative deepening search can guarantee to find the optimal solution. (T/F)

F just the shallowest

A local beam search with k states is the same as running k random restarts hill climbing algorithms in parallel. (T/F)

F no restarts in local beam

There exists a task env in which every agent is rational. (T/F)

T set performance measure to 0

Uniform-cost search behaves like breadth-first search when

edge costs are constant

Depth first search is a special case of best first search when

f(n) = - depth(n)

Breadth first search is a special case of best first search when

f(n) = depth(n)

Uniform cost search is a special case of best first search when

f(n) = g(n)

evaluation function for A*

f(n) = g(n) + h(n)

What is the major difference between the general tree-search and graph-search algorithms?

graph-search has an explored set so it does not expand the same node twice

Uniform cost search is a special case of A* when

h(n) = 0

Local beam search with k =1 results in what algorithm?

hill-climbing search

Genetic algorithm with population size N=1 results in what algorithm?

random walk

define admissible heuristic

when h(n) <= h*(n) for all n, where h*(n) is the true cost to reach goal (never over estimate h)