Machine learning algorithms for prediction of heart disease
| Autor: | Rashmi Ramesh Rachh, Shridhar Allagi, B.K. Shravan |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Heart disease
business.industry Computer science Deep learning media_common.quotation_subject medicine.disease Machine learning computer.software_genre Ensemble learning Field (computer science) Task (project management) Support vector machine Voting medicine Artificial intelligence business computer Algorithm LEAPS media_common |
| DOI: | 10.1016/b978-0-12-821633-0.00010-6 |
| Popis: | The field of healthcare has continuously been under the radar over the last few decades and, as such, has seen advancements by leaps and bounds, especially in the sector of disease prediction and potential data analysis. The quick and timely prediction of diseases is of paramount importance in healthcare systems. The complexity of disease prediction far surpasses the cognitive abilities of the human species. That, combined with the time constraints, make this task an extremely challenging one. The mission of analyzing high-degree medical data concerning all dimensions is monotonous and prone to human errors. This monotonous and tedious task can be efficiently performed by machines equipped with advanced hardware units (such as GPUs). The ability of machines to work tirelessly without error can be harnessed for better accuracy. Recently, Machine learning (ML)/deep Learning (DL) techniques have gained a lot of momentum in the healthcare industry for the fast prediction of diseases with minimal human intervention. Some contributory risk factors such as diabetes, high blood pressure, high cholesterol, etc., make the task of identifying heart disease a difficult one. Owing to this, the heart disease dataset has been chosen for the case study for this chapter. The steps to be followed in using various basic ML algorithms and ensemble ML along with the comparison of results are presented. In our experimentation with base learners, the support vector machine had the highest accuracy of up to 83%. In ensemble learning models, we achieved higher accuracy for weighted average voting of 83%. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|