Fall 2020 MSBC 5030 Quantitative Methods Final Project (R)
Jack Beck, Dria Fabrizio, Jordan Waldroop
This project was assigned as our final project for our Quantitative Methods statistics class. The goal of the project was to perform a regression analysis on a dataset with an identifiable business problem. The choice of dataset was left up to each team. Our team choose a simple dataset of car miles-per-gallon data, which contained a number of cars and their relevant specifications. Using the specifications as features, we created a linear regression model which predicted a car's MPG using those input features.
Link to dataset: http://archive.ics.uci.edu/ml/datasets/Auto+MPG
In order to align with the company’s goal to remain environmentally friendly when purchasing fleet vehicles, an analysis was conducted to determine the most fuel efficient vehicles. Under the assumption that this is based in the 1980’s, with ample availability of vehicles in the dataset, and the price point is not relevant to the scenario, we have determined the specifications for the ideal fleet vehicles for the company.
The key indicator variables are horsepower, weight, and cylinder count. The recommended fleet vehicles will meet the following specifications: 75-150 horsepower, 2000-2500 pounds, and 4 cylinders.
When the recommended specifications were applied to the cars available in the dataset, we found that the 1980 Volkswagen Rabbit (41.5 MPG, 76 HP) and the 1980 Datsun 510 hatchback (37.0 MPG, 92 HP) were the best choices for fleet purchase. Overall, there are 60 vehicles in the dataset that meet the specifications for the ideal fleet vehicle.