Predictive modelling of customer value

Michał Zawal

Diplomová práce

Abstract:

The presented thesis aims to present a way of data analysis, and more precisely, predictive analytics can be applied in many fields of science, knowledge development and the process of gaining a competitive advantage. The main objective of the research is to apply statistical models of supervised machine learning in an attempt to identify the consumer characteristics that have the greatest influence on purchasing decisions. These studies are expected to have a positive effect both on the company using these techniques through appropriate customer targeting, but also on the customers themselves, as the product will be able to reach the interested audience.The impact of the indicators of the prediction model will also be compared to formerly conducted studies of other analysts. Besides merely personal characteristics, the impact of the macroeconomic environment in which decisions were made and the history of consumer behavior will also be assessed. Additionally, before building the model, the customer data set will be well examined in detail using univariate and multivariate methods of statistical analysis. The data will be then cleaned and prepared for the model construction. The final step assume building predictive model and assessing it's performance.

Keywords

Predictive analytics, Machine learning, Statistical analysis

Jazyk práce: angličtina

Datum vytvoření / odevzdání či podání práce: 7. 4. 2020

Obhajoba závěrečné práce

• Vedoucí: Ing. Tomáš Hlavsa, Ph.D.

Česká zemědělská univerzita v Praze

Provozně ekonomická fakulta

Magisterský studijní program / obor:
Systems Engineering and Informatics / Informatics