Breadth First Search applies goal test to each node when it is generated rather than when the node is selected for expansion. Therefore, once the algorithm determines shallowest goal node, it stops the search. By following your explaination, the time complexity of BFS comes around in the order of O( b^ (n+1) ). However, if the nodes were to be tested for goal nodes when they were generated rather than when selected for expansion, time complexity becomes O(b ^ n) since whole layer of nodes at depth n would be expanded before goal was detected. b -- branching factor n -- depth of shallowest goal node.
A couple of drawbacks to bring up. Not guaranteed to find the least costed path from S to a Goal State. Also would use a ton of storage to maintain all the current unexpanded nodes. Back in college we had to solve the canibal missionary problem using DFS, because the storage required would grow exponentially if BFS is used.
If let's say there is only one goal state (G1). Could I accept G1 as the goal state as soon as I add it to the frontier? By the way thank you for your clear explanation!
great explanation!!! could you please upload more videos about Neural networks and probability problems? such as Bayes networks,approximate inference and CNN
I wasted hours at my uni with professors that overcomplicated these search algorithms. You sir saved me big time!
I relate
Breadth First Search applies goal test to each node when it is generated rather than when the node is selected for expansion. Therefore, once the algorithm determines shallowest goal node, it stops the search.
By following your explaination, the time complexity of BFS comes around in the order of O( b^ (n+1) ).
However, if the nodes were to be tested for goal nodes when they were generated rather than when selected for expansion, time complexity becomes O(b ^ n) since whole layer of nodes at depth n would be expanded before goal was detected.
b -- branching factor
n -- depth of shallowest goal node.
Loving this channel, super well explained. Thank you!
I love John's Lecture so helpful when it comes to explaining, I would love to have him as my AI Professor. anyways Thanks Dr John
Congratulations for your explanation! It was the best one for this topic for me!
You are the best profesor ever
the best explanation I found so far...thank you sir
Amazing work brother!
A couple of drawbacks to bring up. Not guaranteed to find the least costed path from S to a Goal State. Also would use a ton of storage to maintain all the current unexpanded nodes. Back in college we had to solve the canibal missionary problem using DFS, because the storage required would grow exponentially if BFS is used.
Wow simple and clear
Thank you Sir
Much appreciated :)
You're very welcome!
Thanks. This is very helpful.
If let's say there is only one goal state (G1). Could I accept G1 as the goal state as soon as I add it to the frontier? By the way thank you for your clear explanation!
Thanks for the explanation sir!
great explanation!!! could you please upload more videos about Neural networks and probability problems? such as Bayes networks,approximate inference and CNN
Excellent explanation
This guy saving my degree
WOW great video!
thankyou sir!!! very well explained
thank you it was great ❤🌹
Thanks
Thank you
Well explained
don't we need to add all the visited letters for that path?
Missing your videos sir
smart fella
thank u
radu cretulescu trece-ma la examen
Wrong! Correct Path is: ABDG1
That's the order the nodes get visited in, but the path is the route that the agent needs to take to get to the goal.