PREDIKSI PENERIMAAN MAHASISWA MENGGUNAKAN NEURAL NETWORK BERBASIS RAPIDMINER PADA DATA GRADUATE ADMISSION

Abstract

This study aims to predict graduate admission outcomes using a Neural Network approach implemented in RapidMiner. The dataset was processed through a series of stages, including data cleaning, normalization, and model training, to ensure optimal learning quality. Model performance was assessed using the Root Mean Square Error (RMSE) metric. The resulting RMSE score of 0.054 indicates a low level of prediction error and demonstrates that the constructed model performs reliably. These findings highlight the potential of Neural Networks as an effective analytical tool for estimating student admission likelihood with higher accuracy and supporting data-driven decision-making in the selection process.