Neural network method is discussed and applied to error decomposition theory of dynamic measurement.
在讨论神经网络方法的基础上，尝试将其应用于动态测量误差分解理论。
Subsequently it establishes error correction models to describe short-term dynamic change and carries on forecast variance decomposition.
在此基础上建立了描述上海城市竞争力短期动态变化的误差修正模型，并进行了预测方差分解。
应用推荐