Visualization of regression models

Abstract

Regression models are used for prediction and forecasting in a wide eld of business areas. To interpret such models users often consider visualization. Due to a missing of simply and precise representation user without sufficient mathematical understandings are not able to understand the inside of such a model. The use case for this thesis is related to the car insurance sector. Nowadays insurances are willing to proceed claims fast. In case of a total damaged car the insurance is currently forced to wait out the offer of the car trader. In average an advertisement of car takes four days. In case of claim processing in comprehensive insurance the insurance could save this time with a prediction. The elaborated artefact should ll up this gap by providing a solution that breaks down the complexity of mathematical model via a graphical representation in analogy to something the user is already familiar with. The development of the new plot has been always with a focus on novice user. According to standard guidelines in the eld of visualization and frameworks in computational statistics the author worked on a new solution in this specific area.The result of this thesis is a new design artefact called the Standardized Novice Plotembedded in an application created with Shiny called Visualization of Regression Model for novice statistical users VRMN.