Tree models provide one way of handling locational considerations. In this assignment we use a standard regression model (OLS) to handle the non-locational variables and see how much trees can explain the part that OLS cannot explain (the OLS residuals). This exercise reinforces standard regression as well as a machine learning technique. Instructions Take the running Louisiana real estate data class_data_state.mat and run an OLS regression with the dependent variable of logprice against the explanatory variables of log(Area), log(Income), BA, Age, and dummies for Sales Year. Take the residuals from this regression and compute the R-squared. Now use the tree model with explanatory variables of Lat and Lon to see how much this model can find signal in the OLS residuals. Specifically, what is the R-squared of the OLS residuals as you change the depth of the tree (or number of nodes)? Use a MATLAB livescript file and export it as a pdf. Include an annotation explaining each output and command