Application of Neural Network Method to Determine Public Satisfaction Level on Pertalite Fuel

Abstract

This research aims to analyze public interest in Pertalite fuel using the Data Mining method, specifically using the Neural Network method. The stages in this research include Data Analysis, Data Preprocessing, Designing Classification Models in Data Mining, Classification Results in Data Mining, Designing Evaluation Models in Data Mining, and Evaluation Results on Data Mining. The classification results show that of the total of 105 community data analyzed, 97 community data showed interest in Pertalite fuel, while only 8 community data showed no interest. The accuracy results obtained were 100%, indicating that the Neural Network method is very suitable and effective in classifying people's interest in Pertalite fuel. The Data Analysis process was carried out to understand and analyze the characteristics of data regarding public interest in Pertalite fuel. Data preprocessing is carried out to clean, transform and integrate data so that it is ready for the classification process. Next, the Designing Classification Models in Data Mining process is carried out to design a classification model using the Neural Network method. Classification Results in Data Mining produces information that the majority of people have an interest in Pertalite fuel. Designing Evaluation Models in Data Mining is carried out to design classification evaluation models, which then produce Evaluation Results on Data Mining which show an accuracy level of 100%. Thus, this research shows that the Neural Network method is very effective in classifying people's interest in Pertalite fuel.