News for hackers. Stuff that matters.
Re:Fizz Buzz in Tensorflow (2016) by Scea91 on Monday February 12, 2018 @08:55AM (#16358524)
Nice. He built an overcomplicated algorithm that is wrong. FizzBuzz was created to also identify this kind of candidates apart from those that can't program at all.

Nevertheless, if I was an interviewer I would have several questions that were unanswered in this iamverysmart writeup:

1) Why did you choose to encode 1 dimensional input as 10 dimensional input with the binary encoding?

2) Do you think that one source of error could be using different underlying distribution for the training and testing dataset? You are basically trying to extrapolate the training dataset.

3) Since this problem can be easily visualized, plot the learned function.

Re:Fizz Buzz in Tensorflow (2016) by Q6T46nT668w6i3m on Monday February 12, 2018 @09:12AM (#16358635)
It’s a joke. :-)

Regarding your second point, it would’ve been less funny if they didn’t overfit. :D