Prediksi Pencapaian Target Kerja Menggunakan Metode Deep Learning dan Data Envelopment Analysis

Abstract

Along with the rapid development of technology, especially in the computer field, several methods have been developed for target setting. Data Envelopment Analysis (DEA) is commonly employed to analyze efficiency levels based on historical data with static targets. Data Envelopment Analysis results in a low level of efficiency against the use of static targets. A new target setting solution is needed to handle dynamic targets.   Based on the need, we propose a method to predict more realistic dynamic targets using Deep Learning Long Short Term Memory (LSTM) approach from the results of the Data Envelopment Analysis (DEA). This study leads to a prediction model with 71.2% average accuracy.