My question is, can we solve [this problem](https://icpc.kattis.com/problems/posterize) using algorithms like `k-means`? After all, all we're trying to do is `minimize the error for each pixel with respect to the nearest cluster/value out of the K values`, right? this idea hit me first before the straightforward dynamic programming solution which passes nicely here [link](https://pastebin.com/nrmju5cS), but when I tried this solution [link](https://pastebin.com/kdj4rXku) I got `WA on test #9` and I have no idea if the error is from my implementation or that the algorithm doesn't even fit here, in which case I'd very much like to know explicitly why?↵
↵
Thanks in advance.
↵
Thanks in advance.
