@lfsim
Lead Data Scientist at mlanalytics.ai – AI in Space researcher – Machine Learning for ESA's Ariel mission (2029) – Formerly at European Space Agency's Advanced Concepts Team and Frontier Development Lab. https://orcid.org/0000-0002-9164-5626
Extreme Learning Machines for Exoplanet Simulations: A Faster, Lightweight Alternative to Deep Learning
https://arxiv.org/pdf/2506.19679
Tara P. A. Tahseen, Luís F. Simões, Kai Hou Yip, Nikolaos Nikolaou, João M. Mendonça, Ingo P. Waldmann.
"We are like butterflies who flutter for a day and think it's forever."
- Carl Sagan, in Cosmos (1980)
Visualisation of our galactic cluster 💫😲💫✨💫🌌💫
NASA
The Ariel team stand smiling together at NeurIPS conference 2024
The Ariel team at NeurIPS conference 2024! 😊
Tomorrow, at #NeurIPS2024, don't miss the session on the "Ariel Data Challenge 2024: Extracting exoplanetary signals from the Ariel Space Telescope".
neurips.cc/virtual/2024...
Meet the winners of the Kaggle competition, and find out the role ML will play in the @arieltelescope.bsky.social
Thanks for creating this.
I'll be going as well.
🚀 Skada v0.4.0 is out!
Skada is an open-source Python library built for domain adaptation (DA), helping machine learning models to adapt to distribution shifts.
Github: github.com/scikit-adapt...
Doc: scikit-adaptation.github.io
DOI: doi.org/10.5281/zeno...
Installation: `pip install skada`
Proud to announce our NeurIPS spotlight, which was in the works for over a year now :) We dig into why decomposing aleatoric and epistemic uncertainty is hard, and what this means for the future of uncertainty quantification.
📖 arxiv.org/abs/2402.19460 🧵1/10
Welcome to Bluesky @arieltelescope.bsky.social ! 👋
That was a great consortium meeting.
Happy to have been able to help with the local organization.
Constructing Impactful Machine Learning Research For Astronomy: Best Practices For Researchers And Reviewers
astrobiology.com/2023/10/cons... #astrobiology #Astrochemistry #astronomy
A high-level summary diagram taken from the slides linked below. It shows the interplay of two main components: a probabilistic model and decision maker or planner.
Probabilistic predictions of an underfitting polynomial classifier on a noisy XOR task and the corresponding under-confident calibration curve.
Probabilistic predictions of an overfitting polynomial classifier and the resulting overconfident calibration curve on the same noisy XOR problem.
Simulation study to show the relative lack of stability of hyperparameter tuning when using hard metrics such as Accuracy or soft yet not probabilistic metrics such as ROC AUC compared to a strictly proper scoring rule such as the log-loss.
I recently shared some of my reflections on how to use probabilistic classifiers for optimal decision-making under uncertainty at @pydataparis.bsky.social 2024.
Here is the recording of the presentation:
www.youtube.com/watch?v=-gYn...
This high resolution model (1.5km) 3D model of Carbon Dioxide shows its sources of emissions and its transport across the globe.
Full details and credits here: svs.gsfc.nasa.gov/5196/
CC: @astrojake.bsky.social @mustaric.bsky.social @maggiebeth.space @nplinnspace.bsky.social @bibianaprinoth.ch @mattkenworthy.bsky.social @donnainfiorino.bsky.social @drjovian.bsky.social @megschwamb.bsky.social @jaynebirkby.bsky.social @astrojaket.bsky.social @niamhk12.bsky.social
This might interest you @cmundell.bsky.social @chrisinbaltimore.bsky.social @sarahkendrew.bsky.social @astrorickman.bsky.social @johndebes.bsky.social @scottwfleming.bsky.social @astrokatie.com @philplait.bsky.social @astrobiology.bsky.social
Could you please help spread the word?
This in-person event, completely free, will be hosted on ESA’s powerful Datalabs platform datalabs.esa.int, offering access to terabytes of data and cutting-edge GPUs.
Small prizes await the top teams who will help push the boundaries of data exploration and discovery!
The hackathon welcomes participants from all backgrounds—no prior experience needed! Connect with experts in both ML and space science, including members involved in the Ariel Space Mission and ESA scientists, for a chance to learn, collaborate, and innovate.
en.wikipedia.org/wiki/ARIEL
Join us for the first #ESA Datalabs Hackathon.
Develop #MachineLearning models to study #Exoplanets, for the #ArielTelescope!
📍 Where: European Space Astronomy Centre (ESA-ESAC), near Madrid 🇪🇸
📅 When: 16-17 January 2025
👉 Register by 15 December 2024: www.ariel-datachallenge.space/esa-datalabs...
Crazy interesting paper in many ways:
1) Voice-enabled GPT-4o conducted 2 hour
interviews of 1,052 people
2) GPT-4o agents were given the transcripts & prompted to simulate the people
3) The agents were given surveys & tasks. They achieved 85% accuracy in simulating interviewees real answers!
This is one really big area of @ec-euclid.bsky.social Legacy Science that absolutely needs human eyes. The very best machine learning gravitational lens finders make lens-candidate lists that are just 10% pure, ie 1 lens comes with 9 false positives. We need your 👀!
🔭🧪
bsky.app/profile/elsa...
