A visual proof that neural nets can compute any function
neuralnetworksanddeeplearning.com
A visual proof that neural nets can compute any function
1–10 of 67 posts
Re: A visual proof that neural nets can compute any function
#2this is really cool
Re: A visual proof that neural nets can compute any function
#3I don't think this always works if the function in question has infinitely many discontinuities...
Re: A visual proof that neural nets can compute any function
#4I don't think this always works if the function in question has infinitely many discontinuities...
What would be an example of a practical task that needs to be modeled with such a function?
Re: A visual proof that neural nets can compute any function
#5Re: A visual proof that neural nets can compute any function
#6I don't think this always works if the function in question has infinitely many discontinuities...
Also a single layer is not enough to approximate functions in higher dimensions.
Re: A visual proof that neural nets can compute any function
#7I don't think this always works if the function in question has infinitely many discontinuities...
You missed this paragraph, which occurs in the first 10% of the article.
The second caveat is that the class of functions which can be approximated in the way described are the continuous functions. If a function is discontinuous, i.e., makes sudden, sharp jumps, then it won't in general be possible to approximate using a neural net.
Re: A visual proof that neural nets can compute any function
#8Re: A visual proof that neural nets can compute any function
#9Re: A visual proof that neural nets can compute any function
#10this is really cool
Isn't it so that all continous functions can be approximated with sums of sine waves too?
That is also cool but without the AI connotations...