IJSECS-4-002


CATEGORIZATION OF GELAM, ACACIA AND TUALANG HONEY ODOR-PROFILE USING K-NEAREST NEIGHBORS

Nurdiyana Zahed1, Muhammad Sharfi Najib1 , Saıful Nizam Tajuddin2
1Faculty of Electrical and Electronic Engineering
University Malaysia Pahang
Pahang, Malaysia
This email address is being protected from spambots. You need JavaScript enabled to view it., This email address is being protected from spambots. You need JavaScript enabled to view it.
2Faculty of Industrial Science and Technology
University Malaysia Pahang
Pahang, Malaysia
This email address is being protected from spambots. You need JavaScript enabled to view it.

ABSTRACT
Honey authenticity refer to honey types is of great importance issue and interest in agriculture. In current research, several documents of specific types of honey have their own usage in medical field. However, it is quite challenging task to classify different types of honey by simply using our naked eye. This work demostrated a successful an electronic nose (E-nose) application as an instrument for identifying odor profile pattern of three common honey in Malaysia (Gelam, Acacia and Tualang honey). The applied E-nose has produced signal for odor measurement in form of numeric resistance (Ω). The data reading have been pre-processed using normalization technique for standardized scale of unique features. Mean features is extracted and boxplot used as the statistical tool to present the data pattern according to three types of honey. Mean features that have been extracted were employed into K-Nearest Neighbors classifier as an input features and evaluated using several splitting ratio. Excellent results were obtained by showing 100% rate of accuracy, sensitivity and specificity of classification from KNN using weigh (k=1), ratio 90:10 and Euclidean distance. The findings confirmed the ability of KNN classifier as intelligent classification to classify different honey types from E-nose calibration. Outperform of other classifier, KNN required less parameter optimization and achieved promising result.


Keywords: Honey; Electronic Nose; Mean feature; Intelligent classification; K-Nearest Neighbors

pdf ico FULL PAPER