6
$\begingroup$

I work in a problem where:

  • The input data is a collection of time series from different sensor modalities, where each sensor comes with different acquisition rates and dynamic ranges
  • The number of potential features vastly exceeds the number of instances (i.e. $p \gg N$)

My question is: Has deep learning been applied on this type of data before?

$\endgroup$
1
  • 1
    $\begingroup$ I have seen random forests of gradient boosted trees applied to dimensionality reduction for that type of data. $\endgroup$ Apr 9, 2014 at 17:23

1 Answer 1

1
$\begingroup$

I asked a very related question (Successful applications of deep learning apart from image classification and NLP) more than two weeks ago and got no answer. My own search turned up only very few deep learning approaches apart from image analysis and natural language processing. Hence, I am quite sure that the answer to your question is: no, deep learning hasn't been applied to such a problem up to now. Go ahead and be the first one ☺

$\endgroup$
1

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.