Electronic Nose Coupled with Chemometrics for Monitoring of Tempeh Fermentation Process

Abstract

Tempeh is a traditional Indonesian food made from fermented soybeans using Rhizopus sp. Tempeh has a large opportunity in the global industrial scale, so monitoring the production process of tempeh is needed. In this paper, the electronic nose is used to study the tempeh fermentation process and to study the stages of this process using linear discriminant analysis (LDA) method. The results of LDA show that there are six stages of the fermentation of tempeh. The stage-1 to the stageS is a process of fungus growth, while the stage-6 is a process of decay. The stages of the tempeh fermentation process are classified using four classification methods: LDA, support vector machine (SVM), k-nearest neighbors (KNN), and random forest (RF). The highest accuracy of classification results obtained RF with an accuracy of 97.33%. The results show good agreement that electronic nose can be used as a standard tool for testing quality of tempeh.