Introducing multiple linear regression
It is time to introduce you to the topic of MLR. In SLR, you will use a single predictor variable. Most business problems deal with outputs dependent on more than two input variables. MLR is the technique used for situations having two or more predictor variables.
In the Using a simple linear regression section, you looked at revenue as a function of the marketing budget. You learned a great deal about the relationship by regressing the revenues variable on the marketing_total variable.
The total amount spent on marketing is the sum of google_adwords, facebook, and twitter marketing expenditures. Using MLR, you can examine the relationship among revenue and some or all of these component budgets.
You will formulate an MLR similar to an SLR, but with more predictor terms. You can think of this formulation as Y regressed on X1 and X2 and so forth. The most common relationship is an additive relationship, and the + operator is used to include multiple variables...
