4.2 Learning Objectives

At the end of this weeks learning, which includes the asynchronous lectures, reading the textbook, this live session, and the homework associated with the concepts, student should be able to

1. Recognize that the conditional expectation function, the CEF, is a the pure-form, best-possible predictor of a target variable given information about other variables.
2. Recall that all other predictors, be they linear predictors, non-linear predictors, branching predictors, or deep learning predictors, are an attempt to approximate the CEF.
3. Produce the conditional expectation function as a predictor, given joint densities of random variables.
4. Appreciate that the best linear predictor, which is a restriction of predictors to include only those that are linear combinations of variables, can produce reasonable predictions, and anticipate that the BLP forms the target of inquiry for regression.