We consider the problem of forecasting net load over a horizon such as one day,\nusing a trailing window of past net load values as well as date and time. We focus\non three variations on this problem: point forecasts, marginal quantile forecasts,\nand generating conditional samples of the future value. We propose a method\nthat relies on linear regression using some custom engineered time-based features\nto capture multiple periodicities, such as daily, weekly, and seasonal, and their\ninteractions. Our proposed …