@jayashrita.bsky.social
Our preprint is now online 🔗https://arxiv.org/html/2509.17937
With @hummerlab.bsky.social, we propose random compression as a quick, non-intuitive way to reduce the input feature space for MD trajectory analysis—while preserving metastable states & kinetics, and yielding more robust results.
With @hummerlab.bsky.social, we propose random compression as a quick, non-intuitive way to reduce the input feature space for MD trajectory analysis—while preserving metastable states & kinetics, and yielding more robust results.
Random functions as data compressors for machine learning of molecular processes
https://arxiv.org/pdf/2509.17937
Jayashrita Debnath, Gerhard Hummer.
https://arxiv.org/pdf/2509.17937
Jayashrita Debnath, Gerhard Hummer.
https://arxiv.org/abs/2509.17937
arXiv abstract link
arxiv.org
September 23, 2025 at 6:21 PM
Our preprint is now online 🔗https://arxiv.org/html/2509.17937
With @hummerlab.bsky.social, we propose random compression as a quick, non-intuitive way to reduce the input feature space for MD trajectory analysis—while preserving metastable states & kinetics, and yielding more robust results.
With @hummerlab.bsky.social, we propose random compression as a quick, non-intuitive way to reduce the input feature space for MD trajectory analysis—while preserving metastable states & kinetics, and yielding more robust results.