Skripsi
Komparasi Algoritme Multilayer Perceptron (MLP) dan Logistic Regression Dengan Seleksi Fitur Particle Swarm Optimization pada Dataset Diabetes
bibliografi: hal. 78
Abstrak
Diabetes mellitus (DM) or called diabetes a metabolic disorder disease is just the characteristic of hyperglycemia that occurs because the pancreas does not produce enough insulin or the body cannot use the insulin produced effectively. At the last estimate of the IDF there are 382 million diabetics in the world by 2013, while in the year 2035 is expected to be increased to 592 million. Estimated from 382 million people, 175 million of them were yet undiagnosed, thus developing progressive threatened to become unwitting complications and without the prevention. Based on the number of diabetics are very big then needed a way to cope with that amount. One of them with techniquesof classification data miningusing machine learning, is solutions are good enough for use. In this study do comparisons of multilayer perceptron (MLP) algorithms and logistic regression with particle swarm optimization in feature selection (PSO) for diagnosing diabetes using datasets the pima indians. From the results of the comparisons of the resulting value accuracy to diagnose diabetes disease of 77.474 percent logistic regression algorithm and 75.651 percent on the multilayer perceptron by using feature selection PSO.
Keywords: Diabetes, Comparative, Algorithm, Logistic Regression, Multilayer Perceptron.
SI0885 | SI 0885 WIB k | UPT. PERPUSTAKAAN PUSAT (Rak 2) | Tersedia - No Loan |
Tidak tersedia versi lain