José Manuel de Frutos

The class for generating [TimeSeries](https://github.com/arrigonialberto86/deepar/blob/master/deepar/dataset/time_series.py) data fails when using DeepAR because it lacks the 'dimension' attribute.

It would be good to have a probabilistic forecasting method like [DeepAr](https://arxiv.org/abs/1704.04110). I have implemented the following package https://github.com/josemanuel22/DeepAR.jl, and it would be necessary to harmonize and correct the errors...

Implement a heavy-tailed estimator

1

It would be good to have a function that implements a Hill estimator, for example. The Hill estimator is particularly useful in extreme value theory for analyzing the tail properties...

adding docu to mean! explain target should not alias with the source

1

fix issue #71

PR that adds [MMD-GAN](https://arxiv.org/pdf/1705.08584.pdf), is implemented for 1D, but we can easily make changes to integrate multiple dimensions. A possible suggestion that I see is to use `KernelFunctions.jl` instead of...

I have followed the download instructions for the Docker from the documentation and I am getting the following error on my Apple `M1 Pro Mac, 13.2.1 (22D68), 16Gb RAM`, when...

Adding Stacked_MNIST #230

2

Adding Stacked_MNIST #230

For Vision Datasets, it would be good to add [stacked-mnist](https://paperswithcode.com/dataset/stacked-mnist) which is an mnist dataset with three channels. It is very useful in the field of GANs for analyzing mode...

adding hillestimator

4

fixing issue #29

Proposal to Implement Code Formatting Standards for Contributor Consistency

1

I haven't found a `.JuliaFormatter.toml` or a `FormatCheck.yml`. I think it's important to have these if other users are going to contribute to the package. I can add them myself....