Statistical Models for Probabilistic Forecasting概率预报的统计模型.pptVIP

AMS Short Course on Probabilistic Forecasting, 9th January 2005 Statistical Models for Probabilistic Forecasting Ian Jolliffe i.t.jolliffe@reading.ac.uk Statistical Models A model can be written in general form as y = g(x) + e. y is something to be predicted (a predictand). It might be a spatial field, but to keep things simple, suppose it is a single quantity at a single location, perhaps temperature, windspeed or probability of precipitation. Statistical Models II y = g(x) + e x denotes one or (often) more predictors – it could be some