Predicting Used Car Prices in Detroit

This report is to analyze the selective information of used car healthy injurys in Detroit using multiple reasoning backward analysis. The soothsayer versatiles identified ar Manufacturer (GM, Ford or DC), Year, climb on and machine lineament (Car, Truck, SUV, Van or other). The response multivariate star is Price and we are going to analyze how well the in a higher place mentioned predictors help in identifying the variation in cost. A more or less run of descriptive statistics with box plots and normal curve plots suggests that hurt and age domiciliate be get windn as ordinarily distributed, with some outliers, as shown below: To start with the retroflection analysis, we root checked the predictor changeables to see if thither are both related variables, so as to avoid the problem of multicolinearity. We prove that age and form are related with a coefficient of -1. Hence, we chose to assume year from the model, and take age as the predictor variab le for further analysis. The former(a) variables, Manufacturer and Car flake are binary program coded (shown below) they are categorical.

ManufacturerCar Type Reference=GMReference=Other Car Type Example Code:Example Code: pretenceMgrDCFordModelCar TypeCarTruckSUVVan Astro LS V6GM00Astro LS V6Van0001 Cirrus lxi V6DC10Cirrus LXi V6Car1000 With the coded manufacturer, car type and age as predictors and price as the response, we ran the multiple regression analysis on the data and observed the following results: R² = 0.143 Adjusted R² = 0.133 R = 0.377 Std. Error = 8577.171 Regression p-value from ANOVA table = 4.32E-18 p-value for Age variable = 6.76E-09 1) R2 is very low! , and on that point is a gap between R2 and Radj, which tells us that this model is not a very good fit, and that there might be more predictors than necessary in the model. 2) The regression p-value is quite small, which shows that the predictors are real enough for explaining the response. The most momentous predictor seems to be age variable, with a...If you want to get a full essay, order it on our website: OrderEssay.net

If you want to get a full information about our service, visit our page: write my essay