Algoritma Backpropagation Metode Powell-Beale Untuk Prediksi Penyakit Stroke

Verdi Yasin (STMIK Jayakarta, Jakarta, Indonesia)
Ifan Junaedi (STMIK Jayakarta, Jakarta, Indonesia)

Brahmana : Jurnal Penerapan Kecerdasan Buatan
Vol 4, No 1A (2022): Edisi Desember (Spesial Issue)

Stroke is a dangerous disease, a stroke is caused by a blockage of blood to the brain or insufficient blood supply to the brain. Stroke can cause brain damage, long term disability and death. The purpose of this study focuses on stroke prediction using the conjugate gradient Powell-Beale backpropagation algorithm. This stroke prediction data is taken from kaggle which consists of 5110 records. The attributes used to predict stroke consist of 7 attributes, namely age, hypertension, heart disease, marital status, average blood sugar, BMI, and smoking status. The results of this study are the prediction of stroke with MSE training and testing = 0. With a 7-2-1 architecture.