Number sequence classification task

Is that possible to use rnn network to classify sequence of increasing numbers like “Increasing” and sequence of decreasing like “Decreasing”?

Or if I have sequences of numbers representing height of mouth between lips of speaking people and non-speaking ones then can I classify whether sequence of numbers represents speaking person or non-speaking or not?

Or recurrent networks cannot handle this??

I saw solving of text semantic classification example but in that case we have limited set of numerical values representing words while I don’t the limited one.

Is that possible to use recurrent net to handle my problem??

I could attach my notebook if it helps…

In NLP tasks you are usually using a mapping between word indices to dense feature tensors using nn.Embedding. An RNN should be able to take your (floating point) features, however I’m not sure which architecture would work best for your use case.

1 Like

thank you for your opinion