Machine learning approach to restaurant numbers: a study to improve prediction accuracy
MetadataShow full item record
Citation (published version)S. Jung, S. Jang. 2019. "Machine Learning Approach to Restaurant Numbers: A Study to Improve Prediction Accuracy." The 24th Annual Graduate Education and Graduate Student Research Conference in Hospitality and Tourism,. 2019-01-03 - 2019-01-05.
While most causal studies try to find the relationship between market size and the number of restaurants, this study focuses on the difference in predicting the number of restaurants by segments, using machine learning classifiers, which are decision tree, support vector regression, and neural network. The study attempts to provide practical implications in predicting the optimal number of restaurants under the assumption that a market bound exist for restaurants. Results show predictability of the number of restaurants to improve greater for lower priced restaurants than higher priced restaurants when franchise and food retail information was included in the model, which imply uneven market power among lower priced to be greater than higher priced restaurants.