Hello!
I want to do data augmentation on a time-series data. I’m aware of time, frequency, time-frequency domain methods but I lack some real life examples of these methods. Can somebody help me on this subject?
Thank you,
Yuhan Chiang
Hello!
I want to do data augmentation on a time-series data. I’m aware of time, frequency, time-frequency domain methods but I lack some real life examples of these methods. Can somebody help me on this subject?
Thank you,
Yuhan Chiang
This seems like an interesting paper on the topic.
I found it by doing an internet search for “data augmentation for time series”.
Thank you! I am currently reading this paper!
I have figured out some methods to apply data augmentation for my implementation, but I also want to learn the best practices to use it in a NN. Should I apply data augmentation as a separate operation or use it in the NN as in the DL specialization course?
I don’t think there is any one best practice.
Should I do data augmentation on only the training set?
Or can I do data augmentation first and then split my train-val sets?
If I put a data augmentation function in my model, how can I track my train-val split?
Thanks,
Yuhan
Augmentation gives you more data.
You can use it anywhere that you need more data.