Uji Mutu Minuman Tradisional Jamu Menggunakan Electronic Nose Dan Electronic Tongue

Larios, Diego (2026) Uji Mutu Minuman Tradisional Jamu Menggunakan Electronic Nose Dan Electronic Tongue. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5022221079-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Jamu merupakan minuman herbal yang banyak dikonsumsi masyarakat Indonesia, tetapi penilaian mutunya masih banyak bergantung pada metode organoleptik yang bersifat subjektif dan kurang konsisten. Penelitian ini mengembangkan sistem uji mutu jamu beras kencur berbasis electronic nose dan electronic tongue yang dipadukan dengan machine learning agar penilaian mutu menjadi lebih objektif dan cepat. Electronic nose memanfaatkan tiga sensor gas semikonduktor untuk menangkap pola senyawa volatil sebagai representasi aroma, sedangkan electronic tongue memanfaatkan tiga elektroda potensiometri untuk menangkap pola respons ionik. Keluaran kedua deret sensor digabungkan melalui dua skema fusi data, yaitu tingkat fitur (mid-level) dan tingkat data (low-level), lalu diklasifikasikan ke dalam tiga kategori mutu, yaitu Baik, Sedang, dan Buruk, menggunakan Support Vector Machine (SVM) dan Random Forest (RF). Pada validasi silang Leave-One-Group-Out, SVM memberikan kinerja terbaik dengan F1-macro sebesar 82,5% pada skema mid-level dan 81,3% pada skema low-level, mengungguli Random Forest. Pada pengujian terhadap 15 sampel baru, skema low-level mencapai akurasi 100% dan paling tahan terhadap variasi antarsesi, sedangkan skema mid-level menurun akibat pergeseran distribusi keluaran sensor. Informasi pembeda mutu terutama berasal dari deret sensor gas, sehingga skema low-level direkomendasikan untuk penerapan lapangan.
===================================================================================================================================
Jamu is a herbal beverage widely consumed in Indonesia, yet its quality assessment still relies largely on organoleptic methods that are subjective and inconsistent. This study develops a quality assessment system for beras kencur jamu based on an electronic nose and an electronic tongue combined with machine learning so that quality evaluation becomes more objective and faster. The electronic nose employs three semiconductor gas sensors to capture volatile compound patterns representing aroma, while the electronic tongue employs three potentiometric electrodes to capture ionic response patterns. The outputs of both sensor arrays are combined through two data fusion schemes, namely feature level (mid-level) and data level (low-level), and then classified into three quality categories, namely Good, Moderate, and Poor, using Support Vector Machine (SVM) and Random Forest (RF). Under Leave-One-Group-Out cross-validation, SVM achieved the best performance with a macro F1-score of 82.5% for the mid-level scheme and 81.3% for the low-level scheme, outperforming Random Forest. In a test on 15 new samples, the low-level scheme reached 100% accuracy and proved most robust to inter-session variation, whereas the mid-level scheme degraded due to distribution shift in the sensor outputs. The discriminative information for quality comes primarily from the gas sensor array, so the low-level scheme is recommended for field deployment.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Jamu Tradisional, Electronic Nose, Electronic Tongue, Uji Mutu, Machine Learning, Traditional Jamu, Electronic Nose, Electronic Tongue, Quality Assessment, Machine Learning
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7871.674 Detectors. Sensors T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7878 Electronic instruments
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User:	Diego Larios
Date Deposited:	27 Jul 2026 06:17
Last Modified:	27 Jul 2026 06:17
URI:	http://repository.its.ac.id/id/eprint/138230

Actions (login required)

View Item