本文从最小二乘法预测问题与正交问题入手，引出投影算子的定义及其 重要性质.在第三章中，我们将最小二乘法预测理论应用在精确线性理性预期 模型的设定与求解中，并且以现值模型、永久收入假说、债券定价模型、凯南 的生产要素需求函数模型为例，对其加以说明.进一步的，我们补充空间完备 性、条件期望等重要结论.最后，我们介绍非精确模型设定的拓展内容.

This article starts with the least squares prediction problem and the orthogonal problem, and derives the definition of the projection operator and its important properties. In Chapter 3, we apply the least squares prediction theory to the setting and solving of the exact linear rational expectation model, and take the present value model, permanent income hypothesis, bond pricing model, and Kenan’s production factor demand function model as examples to illustrate them. Further, we add important conclusions such as spatial completeness and conditional expectations. Finally, we introduce Expanded content of inaccurate model settings.