sigmoid.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A social space for people researching, working with, or just interested in AI!

Server stats:

594
active users

Hi all 👋

I'll be attending in-person and I'm looking for people to meet and chat about a couple of different topics (for details see my replies to this post).

1. Caravan dataset
2. Operational
3. ML-based World/Landsurface models
4. Explainable AI
5. NeuralHydrology ( library)
6. Anything else you are passionate that you want to share with me.

Reach out via reply, direct message or email.

Boosts highly appreciated ❤️

1. Caravan dataset

I strongly believe that is important to advance . If you have access to streamflow data that you are allowed to share but you don't want to spend months preparing a dataset, let's meet and see if contributing to the Caravan community dataset could be an option.

For details about Caravan, see github.com/kratzert/Caravan

GitHubGitHub - kratzert/Caravan: A global community dataset for large-sample hydrologyA global community dataset for large-sample hydrology - GitHub - kratzert/Caravan: A global community dataset for large-sample hydrology
Frederik Kratzert

2. Operational

At Google, we are working towards a flood forecasting system with global coverage. Our models are already operational in several regions/countries of the world and I am happy to chat with people from the same area to exchange experiences and ideas related to operational modeling.

3. ML-based World/Landsurface models

If you are working on machine learning based foundational/landsurface models I would be very happy to have a chat and to learn more about your research!

4. Explainable AI

Trying to understand what ML-based models learn is interesting me for several years now. For example, why is the LSTM so much better in modeling the rainfall-runoff system than any physical/conceptual model, using the same input data? Is there anything we can learn from this model that we can then integrate into conceptual/physical models?

5. NeuralHydrology ( library)

I'm a strong advocate of and software. As such, we open sourced our research library NeuralHydrology many years ago, trying to make our tools easily accessible to the wider community. If you are using NeuralHydrology in your research and have feedback/feature requests that you would like to share, please contact me.

github.com/neuralhydrology/neu

GitHubGitHub - neuralhydrology/neuralhydrology: Python library to train neural networks with a strong focus on hydrological applications.Python library to train neural networks with a strong focus on hydrological applications. - GitHub - neuralhydrology/neuralhydrology: Python library to train neural networks with a strong focus on ...

6. Anything else!

You are passionate about your research and you think that I might be interested, then reach out and let's meet for a coffee/tea/whatever and let me know about your work. I am always happy to hear about other peoples work/ideas!

@kratzert

One of my favorite examples of this is this beautiful paper by Juan Carrasquilla and Roger Melko: arxiv.org/abs/1605.01735

@cr thanks, I'll add that to my reading list