MLR (Multi Linear Regression) refers to a method of forecasting future demands based on historical sales data. This is also known as 'Causal Method' of forecasting. It assumes there is a co-relation between variables (Ex: Advertising & Sales) and by knowing past sales, future advertising you can predict the future sales
Y = A + B(X) where Y=sales & X= Advertising.
You first need to establish the constants in the equation by using past sales & advertising and then use these constants to predict future sales, assuming you know what the advertising budget will be.
A combination of MLR and Univariate models (weights given to each) is known as Composite Profile.
Normally, this MLR / Composite methods are very complex methods and practically very few people use them. (Atleast I have not seen anyone implement this). This is simply because it is very cumbersome to establish a meaningful relationship between variables (like Sales - Advertising in our example) and it also needs business to maintain more data.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.