This lecture covers the mean-square-error (MSE) problem of estimating a random variable from observations of another random variable, focusing on linear regression models. It addresses the challenge of determining the MSE estimator by limiting the mapping function to affine functions of the input variable. The lecture explains the optimal estimate, the linear least-mean-square-error estimator, and the orthogonality principle. It also discusses data fusion methods and the application of linear models in estimating target properties. The concepts are illustrated through scatter diagrams and regression hyperplanes.