A Proficient Hospital Ratings Aware Patient Churn Prediction And Prevention System Using Abg-Fuzzy And Ner-Gfjdkmeans

Abstract

Patient churn in healthcare denotes the rate at which patients stop visiting or seeking care from a hospital. High churn represents dissatisfaction, better alternatives, or accessibility issues. However, the existing works didn’t improve the healthcare churn prediction regarding the hospital ratings along with demographic information, medical history, and clinical factors. Therefore, this paper presents ABG-Fuzzy and NER-GFJDKMEANS-enabled hospital ratings-aware patient churn prediction and prevention system. Initially, the patient review dataset is taken and then pre-processed to improve the data quality. Afterward, from the pre-processed data, NLP features are extracted, and word embedding is done using BERT. Based on the extracted NLP features and word embedding outcomes, polarity identification is performed by utilizing L3STM. In the meantime, patient readmit and hospital ratings datasets are taken and then pre-processed. After that, data balancing by SBMOTE, violin plot generation, feature extraction, feature selection by QBSOA, and classification by L3STM are done. Thereafter, regarding the patient review outcomes, patient readmit, and hospital ratings, the patient chunk is predicted by using ABG-Fuzzy. Next, based on the patient review, the retain strategies are provided for the high and medium patient churn by employing NER-GFJDKMEANS. The results proved that the proposed model achieved a high accuracy of 98.3512%, which was superior to the prevailing methods.