College - Author 1

College of Engineering

Department - Author 1

Industrial and Manufacturing Engineering Department

Degree Name - Author 1

BS in Manufacturing Engineering

College - Author 2

College of Engineering

Department - Author 2

Industrial and Manufacturing Engineering Department

Degree - Author 2

BS in Industrial Engineering

College - Author 3

College of Engineering

Department - Author 3

Industrial and Manufacturing Engineering Department

Degree - Author 3

BS in Industrial Engineering

College - Author 4

College of Engineering

Department - Author 4

Industrial and Manufacturing Engineering Department

Degree - Author 4

BS in Industrial Engineering

Date

6-2022

Primary Advisor

Jill Speece, College of Engineering, Industrial and Manufacturing Engineering Department

Abstract/Summary

After discontinuing their subscription with Shinyapps and relying on a manual forecasting process, Radiology Associates needs a new method to forecast the number and types of scans that will be executed at each site location. Radiology Associates utilizes Quinsite, which incorporates a live link to their database, as a host for all their Tableau dashboards. This project will create an accurate forecasting model utilizing complex forecasting methods to be hosted by Quinsite which is accessible by all management within Radiology Associates. To begin this process, an exponential smoothing model was created in Tableau to solidify dashboard and storyboard design. Additionally, ARIMA and SARIMAX models were built using TabPy and RServe. Using a MASE score each forecasting method was tested. Using the MASE score and feedback from Radiology Associates during a mock forecasting meeting, exponential smoothing was selected as the most accurate to be in the final design. With the final forecasting method selected the dashboard went through several rounds of slight alterations based on feedback from project stakeholders before being officially handed over to Radiology Associates.

Share

COinS