Neural Networks are universal approximators, It is true for picewise continuous functions?

2 ビュー (過去 30 日間)
George Cybenko had prouven that NN are universal approximators for continuous functions (see http://en.wikipedia.org/wiki/Universal_approximation_theorem). Can Neural Network approximate piecewise continuous functions? In particular in MATLAB NN tool box?
I need a response with references.
  2 件のコメント
Walter Roberson
Walter Roberson 2013 年 6 月 8 日
A homework question?
Platon
Platon 2013 年 6 月 11 日
編集済み: Platon 2013 年 6 月 11 日
No not really! But If you have a relevant response I thank you in advance

サインインしてコメントする。

採用された回答

Greg Heath
Greg Heath 2013 年 6 月 8 日
Yes. However, like fourier series, values at finite discontinuities will be the average of the left and right values.
If you don't have a text, my guess is that if you search on something like
neural uniform approximation
you should find something.
Try the comp.ai.neural-nets newsgroup and FAQ before Googling.
Hope this helps.
Thank you for formally accepting my answer
Greg

その他の回答 (0 件)

カテゴリ

Help Center および File ExchangeSequence and Numeric Feature Data Workflows についてさらに検索

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by