Can Neural Networks Crack Sudoku?
github.com
Can Neural Networks Crack Sudoku?
1–10 of 111 posts
Re: Can Neural Networks Crack Sudoku?
#2Sure, but so can decision trees.
Re: Can Neural Networks Crack Sudoku?
#3You can solve sudoku with a SAT solver. You don't need neural networks, this was an assignment in cs 251 at UIC.
Re: Can Neural Networks Crack Sudoku?
#4Every time sudoku comes up on HN I think back to when Ron Jeffries tried to solve it using Test Driven Development ;)
http://ravimohan.blogspot.se/2007/04/learning-from-sudoku-so... - enjoy :D
Re: Can Neural Networks Crack Sudoku?
#5I find it interesting that when the NN fails to complete the whole puzzle, it seems to fail spectacularly (github.com/Kyubyong/sudoku#results) -- that is, there aren't a lot of good-partial attempts (90% or above). Does anyone know why this might be the case? Does the first wrong placement of a number in a gap essentially ruin the rest of the guesses?
Re: Can Neural Networks Crack Sudoku?
#6I'd say that using neural networks to solve sudoku-like puzzles is a bad idea -- a SAT solver, constraints program or integer program would be all quicker to run and quicker to implement.
Re: Can Neural Networks Crack Sudoku?
#7I find it interesting that when the NN fails to complete the whole puzzle, it seems to fail spectacularly (github.com/Kyubyong/sudoku#results) -- that is, there aren't a lot of good-partial attempts (90% or above). Does anyone know why this might be the case? Does the first wrong placement of a number in a gap essentially ruin the rest of the guesses?
Reminds me of Watson's Jeopardy! exhibition match. Watson was right far more often than the humans, but when it was wrong it was way off base.
Re: Can Neural Networks Crack Sudoku?
#8Evidence that DNNs are somewhere between phase 2 and 3 of https://en.wikipedia.org/wiki/Hype_cycle
Re: Can Neural Networks Crack Sudoku?
#9You can solve sudoku with a SAT solver. You don't need neural networks, this was an assignment in cs 251 at UIC.
yes, a NN isn't the best tool for the job. i guess it's just an experiment if one was able to do this with a reasonable rate of success.
Re: Can Neural Networks Crack Sudoku?
#10I find it interesting that when the NN fails to complete the whole puzzle, it seems to fail spectacularly (github.com/Kyubyong/sudoku#results) -- that is, there aren't a lot of good-partial attempts (90% or above). Does anyone know why this might be the case? Does the first wrong placement of a number in a gap essentially ruin the rest of the guesses?
Probably an early-onset snowballing of what you said. When one wrong placement is made, it's pretty hard to recover for a full puzzle.