Google AutoML only supports Image Classification (AutoML vision). While AutoML Tables is for structured data. I need to do time series classification where there is translation invariance in data.
AutoML Tables is not an option as it won't handle translation invariance. AutoML vision can explore convolution layers for handling translation invariance but unfortunately, it asks to upload jpeg images. I cannot feed 1D time series in AutoML vision.
SO how can I use AutoML for 1D time series classification?