Linear regression is to fit a straight line to your data. In its simplest
form, it is to find the slop (a) and the intercept b for fitting your
Y ~ a * X + b.
Here X and Y are one dimesional data arrays. If Y ~ a * X + b is a "good fit",
then X can be used to predict or explain Y.
For details on how to build and evaluate linear regression modes, read
the Electronic Statistics Textbook .
Excel has a standard macro, Analysis Tool Pak (Tools/Add-Ins...), for
simple statistical analysis. The regression macro can be used for
linear regression. The function linest (Insert/Functions...) can also
be used for linear regression analysis.
For an application example, see How to calculate IC50.