Hybrid Method for Evaluating Feature Importance for Predicting Chronic Heart Diseases
Rashid Nasimov, Nigorakhon Nasimova, Bahodir Muminov
Abstract
Predicting the impact of different factors on the patient’s health is as important as diagnosing diseases, especially when monitoring patients with chronic diseases. To perform this by Artificial Intelligence (AI) methods, it is recommended to determine the features importance (FI) of data. There are a number of methods to evaluate FI. However, we can see a big variation in their results which is difficult to interpret. To solve this issue, we proposed new method which aim is minimizing the differences. Furthermore, to demonstrate the effectiveness of the proposed method we used the extracted FIs as weights of the weighted KNN and compared performances.
Topics & Concepts
Computer scienceArtificial intelligenceBig dataFeature (linguistics)Machine learningData miningVariation (astronomy)Pattern recognition (psychology)PhysicsAstrophysicsLinguisticsPhilosophyArtificial Intelligence in Healthcare