Abstract
In today's big data era, a humongous amount of data are collected from various sources. In many cases, these data are incomplete, imprecise, and uncertain. An illustrative example is the OpenStreetMap project, where users all over the world contribute data on a more or less precise and complete level. This research shows whether these data are suited to support management decisions. A real-world ...
Abstract
In today's big data era, a humongous amount of data are collected from various sources. In many cases, these data are incomplete, imprecise, and uncertain. An illustrative example is the OpenStreetMap project, where users all over the world contribute data on a more or less precise and complete level. This research shows whether these data are suited to support management decisions. A real-world example demonstrates the extent to which location decisions of a fast-food restaurant chain can be reproduced using techniques from the field of advanced analytics. The problem deals with classifying potential locations and comparing the predicted locations with the actual ones. The data used for this example are retrieved from the OpenStreetMap project. We find that the OpenStreetMap data are generally suitable for predicting location decisions. However, the choice of the data analytics technique is crucial. In our illustrative example case, boosted trees resulted in the best forecast, thereby outperforming neural networks, classic trees, and logit models.
Altmetric