Uncertainty Quantification in Time Series
Quantile trends in time series data To understand the trend in time series data, we use quantile regression to determine trends. The advantage of this approach is its flexibility in modeling data with conditional functions that may have systematic differences in dispersion, tail behavior, and other covariate features. In adopting this approach, we reduce the d-dimensional nonparameteric regression problems to a series of additive univariate problems. Next, individual quantile curves are then pecified as a linear b-splines.