Hybridization of cognitive computing for food services

Abstract

The application of data mining technology to food services and the restaurant industry has certain social value. By predicting customer traffic and needs, a restaurant can prepare a reasonable amount of meals for customers according to predicted needs which is conducive to improving the dining experience of customers and also improving the quality of food preparation and making the restaurant itself operate more efficiently. In recent years, we have seen the use of collaborative robots for use in the fast food industry. In Asia and more specifically in Japan, we have seen many fast-food chains implement the use of robots to better serve their customers. By studying the linear regression algorithm and the random forest algorithm, this paper proposes a new interwoven novel fusion approach of combining both algorithms and applies the new model to restaurant data to assist in the prediction of customer traffic in the restaurant industry. This predictive algorithm using cognitive techniques can assist these newly place robots in the food industry better serve their client base and in doing so make the industry more efficient. Experimental, comparison, and analysis are reported in the paper. The error rate of the fusion solution is reduced by approximately 5.503% compared with the linear regression algorithm and is approximately 3.719% lower than the error rate of the random forest algorithm. Results show that the new fusion algorithm can achieve better prediction results of customer traffic prediction for the restaurant industry. Furthermore, we also provide a new take on the application of data mining technology in the restaurant industry itself.