BotClean Stochastic
-
-
dheeraj 10 years ago These are learning based challenges so, the changes needed for subsequent challenges are minimal. That's the whole point of the exercise.
These challenges teaches you about different environments in which a rational agent of AI is put into and how it should behave.
-
yashdeora98294 3 years ago here is problem solution in java python and c++ programming. https://www.gyangav.com/2022/03/hackerrank-botclean-stochastic-problem-solution.html
-
mohammed_abdali1 3 years ago But what was I supposed to have had learned? The original one I programmed as "Check to see if standing on dirty floor. If so, clean it and quit. Otherwise start looking at the top left corner and read the grid like you would a book (the normal method of reading a matrix, that is). As soon as you encounter a dirty floor, make note of it and stop reading the grid. If your robot is too low or too high, move one step up or down as needed, and then exit. However, if you're already at the right level, then move either left or right as needed, and exit."
This exact logic works for "what if I generate a new dirty floor after you clean the first one?" I guess the only technicality is that I have 200 steps instead of infinity (I think I had infinite in the first challenge), but considering I got full points, it looks like that constraint doesn't matter...
Unless there was an easier (less complex) solution that was expected for the first challenge. But I mean... what's less complex than the brute force method of "try to find the lowest index of a 1d array that has dirt, and then try to move towards it"?
-
-
alan_davies 9 years ago I submitted the same code too; it's just a simplification of botclean.
-
facundoq 9 years ago Same here. I also feel these exercises should be reversed. The first one is much harder, as the optimum involves calculating the shortest path through the grid that passes through all the dirt positions.
-
omkarbhale001 4 years ago No, even if you(and I) did the previous challange in optimised way, I think the sequence did not matter for previous challenge.
-
-
-
simonesturniolo 8 years ago Same here. The only real difference is in efficiency. The stochastic one could be made to run faster in terms of execution time, but not in less moves. If time was considered too in the scoring then it'd be a bit more interesting.
-
ayusmittal 8 years ago You have a very good point about time execution. The system should also generate the execution time and then compare and give ranking to results. It would be a great way to judge.
-
-
scott_mankowitz 8 years ago It's worse than that: the only change I made was changing the function from next_move to nextMove. It seems sloppy on the part of the puzzle designers.
Also, I agree with the others, this problem is easier than the last, and could be used to provide a foundation for it, instead of vice-versa.
-
martink_kadarman 6 years ago I used same code too even it works. i didn't get any point somehow.
-
jonibeknorboev 5 years ago don't worry, if your code didn't behave the same ways as it did for simplified version of the problem then you'd get that point. If your code behave normal in different invironments then probably your model is not bahing abnormal. The point of AI
-
-
tyce_brown 4 years ago Same here. I feel like the BotClean Stochastic was useless, as all I did was copy and paste my code from BotClean.
-
I used the same code for botclean and botcleanr. It's practically the same, except regular one allows for a higher score if more accurate, still I felt it kind of pointless that I did not need a single change.
Add Reply Preview cancel