In this analysis, we are going to use a linear multiple regression. Why is this the most appropriate model for this data set?

Is demand for economy cars seasonal? In other words, does demand differ between the $4$ seasons? To test this, run a Group ANOVA with EconContracts as the DV and season as the predictor.

a$.$ Are there significant differences between groups?

b$.$ Next, look at the means by season by running an Aggregate operator $($aggregate EconContract averages by Season$).$ In which season$($s$)$ is demand the highest?

Run a regression model that includes your choice of dependent variable and include all of the relevant predictor variables $($please be careful to not include other "dependent variables" as predictors $$ think... what are things we would want to predict and what are things that we could use to predict that thing?$).$ DO NOT include the week number variable in the model, and DO NOT include a "feature selection" in your Regression model $($we want to look at all variables before making a more parsimonious model$).$ You will also need to add a "Nominal to Numeric" operator for the season variable $($this will dummy code the variable$).$

a$.$ How would you describe the fit$/$accuracy of the model you created? How did you come to this conclusion?

b$.$ Based on this model, which variables predict demand for economy cars? Discuss the strength and direction of each effect.

c$.$ Pick two significant predictor variables and provide a detailed interpretation of how you can interpret the regression coefficient for that variable.

d$.$ Which variables do not predict demand? How might this inform strategy?

e$.$ In this full regression model, none of the seasons predict economy contracts. Why might this be the case if the ANOVA in Q$2($above$)$ revealed that season did predict the DV$?$

f$.$ Based on your answer to Q$3$d $($above$),$ how would you construct the model moving forward to ensure you are maximizing accuracy while minimizing co$-$linearity?

Rerun multiple regression models with each of the other potential dependent variables $($upgrades$,$ total contracts, contract days, and average contract length$).$

a$.$ Do any of these other models perform any better? In other words; is our predictive ability better for any of the other outcome variables? If so$,$ which one$($s$)?$

The car rental agency wants to set up a machine learning model in order to generate optimized, real$-$time demand estimates for economy cars. Based on your findings, which variables should they incorporate into to this model?

For your assignment

Answer all the questions above, in the order that they are asked.

Format the assignment to look like a business report, using headings as appropriate.

Start with a brief overview of the report. What questions should the reader expect to be answered?

Include visualizations to ensure your findings are clearly communicated.