In the conventional approach to time series analysis, after we have identified and removed trends and seasonality, we assume that the residuals thus formed are a stationary time series. We then utilise particular methods to ascertain whether this residual series is the realisation of a purely random process, or if it contains serial correlation of some nature.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Box–Jenkins Models

  • John Boland

摘要

In the conventional approach to time series analysis, after we have identified and removed trends and seasonality, we assume that the residuals thus formed are a stationary time series. We then utilise particular methods to ascertain whether this residual series is the realisation of a purely random process, or if it contains serial correlation of some nature.