EXCON, a supervised embedding method that regresses multivariate time series to per-class extreme prototype vectors, reports the highest TSS (0.71) among compared baselines on the SWAN-SF solar flare benchmark, alongside weak positive-class skill (F1 = 0.16, GS = 0.02).
A time series classification-based approach for solar flare prediction,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
EXCON: Extreme Instance-based Contrastive Representation Learning of Severely Imbalanced Multivariate Time Series for Solar Flare Prediction
EXCON, a supervised embedding method that regresses multivariate time series to per-class extreme prototype vectors, reports the highest TSS (0.71) among compared baselines on the SWAN-SF solar flare benchmark, alongside weak positive-class skill (F1 = 0.16, GS = 0.02).