Valuation of Secruity

We have looked at break-even calculations, scatter plots for estimating demand information, and trends in revenue so far. In each of these, we simplified the analysis to a straight line. While this is useful, sometimes these data are more complex. They may be a curved line, or instead of only one item for the slope, such as variable costs, there may be several factors involved in the estimate. So demand, for example, may be based on price, convenience, or quality. How does a linear regression allow us to better estimate trends, costs, and other factors in complex situations?