Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 3

Uninformed Search Strategies: Breath-First & Uniform Cost

Breadth-First Search
Background of the algorithm
Definition
BFS is an uninformed search algorithm that systematically traverses a graph
by expanding the shallowest unexpanded node first using a FIFO queue.
Main Proponent and History
Developed in the late 1950s and early 1960s by computer scientists working in
the field of artificial intelligence like Allen Newell, Herbert Simon, and Edgar F.
Codd.

Pseudocode
Step 1: Initialize empty frontier queue Q.
Step 2: Add the start node to Q.
Loop until Q empty:
Step 3: Remove the first node N from Q.
Step 4: Check if N is goal, return path if yes.
Step 5: Expand N, add resulting nodes to Q.
Limitations
Exponential space complexity, slow compared to informed searches, can get
stuck in large cyclic graphs.

Research Utilizing BFS


Research 1:
Title/Author
Breadth-First Search on GPUs (Harish and Narayanan, 2007)
Abstract
Implemented parallel BFS on GPU, achieves speedups of 13-60x over CPU
algorithm.
Research 2:
Title/Author
Parallel Best-First Search for Monte Carlo Tree Search (Sabharwal et al. 2012)
Abstract
Presents a parallel BFS-based Monte Carlo tree search method for game
domains like Go.
My Personal Reflection
BFS was one of the first search algorithms I learned in AI. While simple, it
provides optimality guarantees that advanced heuristics cannot. The key
weakness is space complexity, leading to research into more focused search
techniques.

Uniform Cost Search


Background of the Algorithm
Definition
UCS is an informed search algorithm that expands nodes based on path cost,
using a priority queue ordered by cumulative cost.
Main Proponent and History
Primarily based on Edsger Dijkstra's 1959 algorithm for finding shortest paths
on graphs.

Pseudocode
Step 1: Initialize PQ frontier ordered by path cost.
Step 2: Add start node to PQ with cost 0.
Loop until Q empty:
Step 3: Pop cheapest node N from PQ.
Step 4: Check if N is goal, return path if yes.
Step 5: Expand N, add resulting nodes to PQ.
Limitations
Shares space complexity issue with BFS, slower than algorithms using
admissible heuristics.

Research Utilizing UCS


Research 1:
Title/Author
Anytime Dynamic A* with Uniform Cost Search (Koenig et al, 2007).
Abstract
Presents a variant of A* that uses UCS to efficiently repair solutions anytime.
Research 2:
Title/Author
Uniform Cost Search for Obstacle Navigation (Dolev et al, 2012).
Abstract
Applies UCS on grid for unmanned vehicle navigation with obstacle avoidance.
My Personal Reflection
I appreciate UCS for blending the cost optimization of informed searches with
BFS's simplicity and optimality. The expansive memory requirements inspire
integration with heuristics like we see in A*.

You might also like