부산대학교 도서관

The developing nexus between Internet of Things (IoT) technologies, healthcare, as well as analytics for prediction is the subject of this study. A deductive method is used to investigate how IoT-generated medical data and group learning approaches might be combined for predicting diseases. The study is guided by a descriptive research approach that emphasizes the necessity to thoroughly evaluate any possible overlap between these categories. This investigation's foundation is secondary data collecting from several IoT-enabled healthcare device sources. Studying the effects of picking features overall engineering, researchers found that doing so significantly increased the accuracy of models by 7% and precision by 8%. The evident superiority using ensemble methods is demonstrated by comparison with conventional models, having Random Forest surpassing logarithms approximately 15% in frequency. The study also assesses each model's efficiency of resources and scaling, emphasizing how practically applicable they are to high-throughput, practical applications. The results demonstrate the intriguing possibility of ensemble-based informatics for predicting illnesses using data provided by the Internet of Things. The report makes recommendations for further research, such as investigating cutting-edge ensemble approaches and incorporating other data modalities. To further improve and validate the suggested approach, undertaking real-world validation tests and incorporating methods for interpretation are advised. The ultimate goal of this research is to transform illness management overall patient outcomes by advancing medical predictive analytics.