Prediction of Powdered Coffee Brands Based on Aroma Using Electronic Nose and Artificial Neural Networks

Authors

  • Imam Sofi’i Jurusan Teknologi Pertanian, Politeknik Negeri Lampung
  • Zainal Arifin Jurusan Teknologi Pertanian, Politeknik Negeri Lampung
  • Harmen Harmen Jurusan Teknologi Pertanian/Politeknik Negeri Lampung

DOI:

https://doi.org/10.25181/icoaas.v3i3.2864

Abstract

The way to find out the brand of coffee powder on the market is by looking at the packaging. If you don't know the brand, you can guess by smelling it. The purpose of this study was to predict the brand of coffee powder based on aroma using an electronic nose and artificial neural network (ANN). The method used is to take samples of 3 brands of coffee powder on the market and detect the aroma using an electronic nose. The electronic nose used has 6 sensors. The sensor reading value is used as an artificial neural network (ANN) input. The output of the artificial neural network is the brand of coffee powder. The data used are 240 data, namely 180 data for ANN training and 60 data for validation. The validation results show that the highest accuracy of ANN in predicting the brand of coffee powder is 67.78% in ANN training of 50 thousand iterations.

Downloads

Download data is not yet available.

Published

2023-03-14