RSM412H1 Lecture Notes - Lecture 2: Supervised Learning, Dependent And Independent Variables, Statistical Hypothesis Testing
Document Summary
Both input and desired output data are provided. Training dataset is used to train the machine. Machine is trained to classify something into some class: before doing any modelling, should look at what data looks like first. Main strategy of business defines its analytical choices. Suppose we have an advertising data set, containing advertising budgets for tv, radio, and newspaper, along with total sales for 200 different markets. Can use analysis to determine how to control advertising budget for each medium to maximize sales. Which media generate the biggest boost in sales. How much will sales increase given an increase in tv advertising. How strong is the relationship between advertising budget and sales. Variation does not depend on that of another b. Model: transformation engine that helps express dependent variables as function of independent variables. A weakness of statistical learning is the reliance on the model. Parameters are ingredients added to the model for estimating output.
