Efficient Unsupervised State Detection for Multivariate Time Series
The author proposes E2Usd, an efficient model for unsupervised state detection in multivariate time series, addressing challenges of computational overhead and false negatives through innovative techniques.
