A systematic framework for predictive modeling is proposed, involving data collection, feature subset selection, and classification to identify the best model based on ev...
Abstract:
A company’s survival in the current competitive market hinges on its ability to not only meet but exceed customer expectations, as customers are invaluable assets. Patien...Show MoreMetadata
Abstract:
A company’s survival in the current competitive market hinges on its ability to not only meet but exceed customer expectations, as customers are invaluable assets. Patient satisfaction is crucial in the healthcare sector, directly influencing whether patients will return to a hospital or recommend it to others. This study uses advanced data mining techniques to accurately estimate and predict patients’ likelihood of returning for future appointments by assessing their satisfaction levels. In addition to feature selection models such as Random Forest, Genetic Algorithm, and Lasso Regression, the study employs various methods, including Neural Networks, Support Vector Machines, Decision Trees, k-Nearest Neighbors, Rule-based systems, and Naive Bayes algorithms. The analysis of the results indicates that while the Neural Network model shows superior prediction accuracy, the Lasso Regression method is efficient in identifying relevant features. By integrating AI approaches and thoroughly examining satisfaction ratings in the Iranian healthcare industry, this research makes a significant contribution. Moreover, the findings demonstrate that the Artificial Neural Network model best fits the predictive model and offers the highest reliability. This study aims to forecast patient satisfaction in the healthcare industry and develop a strategic roadmap for hospitals, thereby expanding the knowledge of machine learning methods for predicting customer satisfaction.
A systematic framework for predictive modeling is proposed, involving data collection, feature subset selection, and classification to identify the best model based on ev...
Published in: IEEE Access ( Volume: 13)