3
$\begingroup$

When using CNN on non-image(times series) data prediction, what are some constraints or things to look out for as compared to image data?

To be more precise, I notice there are different types of layers in a CNN model as shown below, which seem to be particularly designed for image data, are they applicable to Non-Image Data(times series)?

A convolutional layer that extracts features from a source image. Convolution helps with blurring, sharpening, edge detection, noise reduction, or other operations that can help the machine to learn specific characteristics of an image.

A pooling layer that reduces the image dimensionality without losing important features or patterns.

A fully connected layer also known as the dense layer, in which the results of the convolutional layers are fed through one or more neural layers to generate a prediction.

$\endgroup$
  • $\begingroup$ One thing I can tell, is that well on translation invariant data like images, whereas time series data is not translation invariant, so try to transform it to translation invariant. $\endgroup$ – DuttaA Nov 29 '19 at 2:41
  • $\begingroup$ @DuttaA Thanks for the advice. Could you please recommend some methods as to how I could transform time series data to translation invariant? $\endgroup$ – nilsinelabore Nov 29 '19 at 2:44
3
$\begingroup$

Usually you need to ensure that your convolutions are causal, meaning that there is no information leakage from the future into the past. You could start by looking at this paper which compares Temporal Convolutional Networks (TCN) with vanilla RNNs models.

| improve this answer | |
$\endgroup$
2
$\begingroup$

You can use CNN in timeseries data. Convolutional Recurrent Neural Network(RCNN) is one of the examples. Convolutional layers basically extract feature from image, It is not related to time series data passing, Neither of them you mention on the question. CNN therefore use some recurrent concept to improve their prediction such as in ResNet, Highway Networks, DenseNet but the all are within single datapoint reasoning. You can go through these concepts to improve your intuitions.

| improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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