Shuai Zhang

Generative time series models can now produce multiple plausible future trajectories as fast as quantile models query a single point, unlocking accurate dependency modeling for critical applications.