Comparison overview methods
Please find an overview of the methods submitted below. If you want to propose a another comparison category, please comment below. Also for submitting teams to correct my comparison categorization answers.
Project | User | Submission tag | RPSS | Visibility | programming language | opportunistic | software | methods | features | CPU vs GPU | |
---|---|---|---|---|---|---|---|---|---|---|---|
1 | s2s-ai-challenge-template | jordan.gierschendorf1 | submission-methods-7 | 0.0459475 | python | yes | pytorch, hydra | climatology, EMOS, ECMWF CNN and CNN weighting | t2m, tp, orography, week, additionally for ECMWF | GPU | |
2 | s2s-ai-challenge-BSC | lluis.palma | submission-ML_models | 0.0288718 | python | yes | sklearn | climatology, raw ECWMF, LogisticRegression, RandomForest | t2m, tp | CPU | |
3 | s2s-ai-challenge-uconn | adam.bienkowski | submission-tmp3-0.0.1 | 0.00620681 | python | no | sklearn | RandomForestClassifier | t2m, tp | CPU | |
4 | s2s-ai-challenge-kit-eth-ubern | horatn | submission_local_CNN_final | 0.00236371 | python | no | tensorflow | CNN | t2m, tp | CPU | |
5 | s2s-ai-challenge-template | damien.specq | submission-damien-specq | 0.000467425 | R | no | script | Bayesian statistical-dynamical post-processing | t2m, tp | CPU | |
6 | s2s-ai-challenge-ECMWF-internal-testers | chantry.m.j | submission-pplnn-0.0.1 | -0.0301027 | python | no | tensorflow | CNN, resnet-CNN or Unet | t2m, tp, sp, sst, sm20, tcw, ttr | CPU | |
7 | s2s-ai-challenge-alpin | julius.polz | submission-resubmit-alpine-0.0.1 | -0.597942 | python | no | tensorflow | Unet | t2m, tp | CPU | |
8 | s2s-ai-challenge-ylabaiers | deepkneko | submission_last | -0.756598 | python | no | pytorch | CNN | t2m, tp | GPU | |
9 | s2s-ai-challenge-kjhall01 | kjhall01 | submission-kyle_nachi_poelm-0.0.6 | -0.9918 | python | no | XCast using sklearn | Probabilistic Output Extreme Learning Machine (POELM) | t2m, tp | CPU |
Edited by Aaron Spring