0

I am working on a school project which requires us to generate a random digraph with weighted edges and perform a number of graph traversal techniques to find one of the target nodes (marked in green) from the starting node (marked in red). A sample graph can be seen below: Randomly Generated Digraph

I've managed to implement the requested uninformed graph search algorithms (Depth-First Search, Breadth-First Search, Iterative Deepening Search and Uniform Cost Search). However, we were also asked to perform A* (A-star) search on this graph and I am struggling to find an admissable heuristic since positional data is required for the Manhattan or Euclidian distances.

Is there a valid heuristic which can be used or is it the case of not having enough information?

Matthew Darmanin
  • 1
  • 1
  • 1

1 Answers1

0

An acceptable heuristic is that the cost is some constant - such as 0.

This will turn A* search into Uniform Cost Search.

btilly
  • 43,296
  • 3
  • 59
  • 88
  • Yes, but having already implemented Uniform Cost Search, this would defeat the point if implementing A* search separately no? – Matthew Darmanin Apr 12 '22 at 18:59
  • @MatthewDarmanin If the point is learning, then no. Also, you can generate randomly in a way that there is a heuristic. For example generate a random graph, place the nodes randomly on the plane, then add Euclidian distance to all edge weights. Now A* search can shine! – btilly Apr 12 '22 at 20:25