IMPLEMENTASI DATA MINING UNTUK PREDIKSI MASA PAKAI KWH METER DENGAN METODE NEURAL NETWORK (STUDI KASUS PT PLN(PERSERO) RAYON PRABUMULIH)
Daftar Isi:
- PT PLN(Persero) is a government-owned BUMN company that serves as electricity provider. One of the divisions at PT PLN (Persero) is the Energy Transaction division which focuses on the sale of electrical energy and minimizes the amount of energy losses that can cause disadvantages. One of the things that affects energy losses is the damaged of kWh meter. This research aims to predict the useful life of kWh meters and it is expected that the results of these predictions will provide useful information to do what steps need to be taken, and can support the achievement of key performance indicators from the energy transaction division of PT PLN (Persero) Rayon Prabumulih. Data used in this research is data replacement of customer’s kwh meter which amounted to 3911 data counted for 8 months in 2017. This research was conducted using neural network method with backpropagation algorithm, the results of the research were classified into 2 classes, above 10 years or under 10 years, then evaluated using precision and recall so that the presicion results were 99.37% and the recall was 29.70%.