An intelligent method of detecting and classifying hotels for an efficient choice
Résumé
The hotel industry attaches a great deal of importance to customer satisfaction and loyalty by offering quality services. However, given the security situation in Burkina Faso, the proximity of administrative and public services, health services and the quality of access roads have become other essential criteria in the choice of accommodation sites. Given that manual analysis of such a large number of criteria and the volume of data generated proves utopian, the search for solutions to aid decision-making is imperative.
To address this concern, we opted for a comparative study of machine learning and multi-criteria optimisation for better prediction of customer choice. We also carried out a comparative study of different classification techniques to determine the best approach.
We then used the correlation matrix to determine the coefficients applied in the Pareto principle.
The different approaches enabled us to confirm that the Extra Trees Regressor gave a better prediction of the choice of hoteliers. We then determined the coefficients of the endogenous criteria in our study using a correlation matrix. From the application of the Pareto principle, we can affirm that: security, pharmacies and health centres are the determining factors in the choice of hotel complexes.
Origine : Fichiers produits par l'(les) auteur(s)