A Predictive Model for Hypertension Diagnosis Using Machine Learning Techniques
| Autor: | Muhammad Ali Javed Tengnah, Soulakshmee D. Nagowah, Raginee Sooklall |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Gestational hypertension
business.industry Computer science Decision tree Machine learning computer.software_genre medicine.disease World health Stress level C4.5 algorithm Multilayer perceptron medicine Artificial intelligence Hypertension diagnosis business computer Back propagation artificial neural network |
| DOI: | 10.1016/b978-0-12-816948-3.00009-x |
| Popis: | Hypertension, a non-communicable disease, is considered to be one of the biggest contributors to the global death rate and according to the latest data published in May 2014 by the World Health Organization (WHO), death rate by hypertension has reached up to 6.39%. This chapter presents a predictive model for diagnosing hypertension. The predictive model is composed of the Hypertension Diagnosis model (HD), Psychological and Stress Level model (PSL) and the Pregnancy Hypertension (PH) model. The models have been implemented using machine learning techniques. The HD model has been implemented using the back propagation Artificial Neural Network (ANN) Multilayer Perceptron algorithm and the other two models have been implemented using the Decision Tree J48 algorithm. The PSL model works side by side with the HD model, whereby the stress level gained from the PSL model is used as input to the HD model. The PH model comes into play when the user is a pregnant woman and has been diagnosed as hypertensive. The three models have yielded the desired accuracy during their training and testing, with the HD model having 98.79%, PSL model having 91.3% and PH model having 100% accuracy. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|