Uninformed Search(Week 3) Flashcards
Uninformed Search Algorithm?
-Blind Search: It doesnot have any additional information about state or search space. It only knows how to traverse through
- No cost consideration
- It might take up more time
types of uninformed search algorithms?
- Breadth first Search
- Depth first Search
- Uniform Cost Search
What properties of search algorithm are considered?
- Optimal: Provides least cost path
- Complete: if it provides any solution path
- Time Complexity: Maximum amount of time needed to traverse through.
- Space Complexity: Least amount of space needed to traverse
Depth first search properties
- Expands the deepest node first
- Starts from leftmost deepest node
- Time taken: O(b^m) where m is maximum depth and is finite
- Space taken: O(bm)
- Can give complete solution
Is DFS complete or optimal? How?
Its complete if m (i.e. maximum depth) is finite.
It may not be optimal because it always finds the left-most solution,regardless of depth or cost.
What are Breadth first search properties?
- Expands the shallowest node first
- Explores all nodes above shallowest node
- Time taken is O(b^s) where s is depth of shallowest solution
- Space is also O(b^s)
- Can be complete or optimal
- It finds the shortest path in terms of number of actions, it might jot always be the least-cost path
Is BFS complete or optimal?
It can be both. Its complete because s (i.e. number of solutions) is finite and its optimal if costs are all 1.
Difference between DFS & BFS
DFS
- Expands deepest node first
- Better in space complexity (b*m)
- For deepest goal
BFS
- Expands tier by tier
- Exponential space complexity (b^m)
- For shallowest goal.
What is Space Complexity & Time Complexity?
Space Complexity: Least amount of space needed by search algorithm to traverse through
Time Complexity: Maximum amount of time needed to traverse
What fringe strategies/ Priority Queues are used for each search algorithm ?
Priority queue is dependent on environment and what is required more
For DFS: LIFO i.e. Last In First Out
For BFS: FIFO i.e. First In First Out (Shallowest)
For UCS: Cost
What are Uniform Cost Search properties?
- Expands cheapest node first
- Space & time complexity stays the same because solution depends on cost
- It provides both optimal and complete solution
Issues of UCS?
- Explores options in every direction
- No information about goal location
- Time complexity is higher