Simulation of Detection of Purchasing Behaviors of People Using Supervised Quantum Machine Learning Based on Continous-Variable Model

In this study, purchasing behavior of people is examined by supervised quantum machine learning based on continuous-variable model. In this context, sample data is taken from the cloud environment. These data is provided as homogenous separation as 75% training and 25% test data. Separated test data was not used in the training process, attention was paid to the healthy implementation of the learning process. Then, the normalization process was performed to ensure the consistency between the independent variables in this data. Thus, the data ready for the learning process are used in the supervised quantum machine learning algorithm performed on variational circuit based continuous variable model. In addition, these data are simulated with the classical support vector machine learning algorithm. Confusion matrices and receiver operating characteristic (ROC) curves for both quantum and classical machine learning algorithms were obtained. Finally, by entering sample values except the test data in the data set, the results are displayed textually and visually. Based on the obtained results, it was determined that quantum machine learning based on continuous variable model is more sensitive. The source codes related to algorithms are found in Github