Aggregation of Experts
摘要
Online robust aggregation of a finite set of forecasts (machine learning method, statistical model, physical model, human expertise, …) is a popular method, well known to perform well on various forecasting applications: meteorological, pollution, finance, and electricity load forecasting. No stochastic assumptions are made on the data generative process, and the theoretical developments are based on the theory of individual sequences where the goal is to derive aggregation strategies achieving good forecasting results on all sequences of observations.