KLASIFIKASI MINYAK GORENG MENGGUNAKAN ELECTRONIC NOSE BERBASIS SENSOR MQ DAN PEMBELAJARAN MESIN

CHAIRUNISA NURUL SADIAH, . (2025) KLASIFIKASI MINYAK GORENG MENGGUNAKAN ELECTRONIC NOSE BERBASIS SENSOR MQ DAN PEMBELAJARAN MESIN. Sarjana thesis, UNIVERSITAS NEGERI JAKARTA.

	Text File Cover.pdf Download (1MB)
	Text File BAB 1.pdf Download (266kB)
	Text File BAB 2 .pdf Restricted to Registered users only Download (787kB) | Request a copy
	Text File BAB 3.pdf Restricted to Registered users only Download (491kB) | Request a copy
	Text File BAB 4.pdf Restricted to Registered users only Download (820kB) | Request a copy
	Text File BAB 5.pdf Restricted to Registered users only Download (220kB) | Request a copy
	Text File Daftar Pustaka.pdf Download (267kB)
	Text File Lampiran.pdf Restricted to Registered users only Download (187kB) | Request a copy

Abstract

Penurunan kualitas minyak goreng akibat penggunaan berulang berisiko terhadap kesehatan, namun sulit dibedakan secara kasat mata. Penelitian ini bertujuan untuk mengklasifikasikan minyak goreng baru dan terpakai menggunakan sistem Electronic Nose (E-Nose) berbasis delapan sensor gas MQ. Sampel terdiri atas minyak baru serta minyak terpakai setelah digunakan 1x, 3x, 5x, dan 7x. Data sensor diolah melalui tahap pra-pemrosesan menggunakan filter Savitzky-Golay, lalu diekstraksi fiturnya (maksimum, rata-rata, dan area bawah kurva). Seleksi fitur dilakukan dengan metode Mutual Information. Selanjutnya, data diklasifikasikan menjadi dua kelas (baru dan terpakai) menggunakan empat algoritma machine learning yaitu Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Gradient Boosting, dan Random Forest. Evaluasi model dilakukan menggunakan Leave-One-Out Cross Validation (LOOCV) dan confusion matrix. Hasil menunjukkan bahwa model Random Forest mencapai akurasi tertinggi hingga 100% dalam pengujian, dengan performa stabil meskipun jumlah fitur dikurangi. Temuan ini menunjukkan bahwa kombinasi E-Nose dan machine learning efektif untuk membedakan minyak goreng baru dan terpakai secara cepat dan non-destruktif. ***** The degradation of cooking oil quality due to repeated use poses health risks, yet it is difficult to distinguish visually. This study aims to classify fresh and used cooking oil using an Electronic Nose (E-Nose) system based on eight MQ gas sensors. The samples consisted of fresh oil and used oil after 1, 3, 5, and 7 frying cycles. Sensor data were preprocessed using the Savitzky-Golay filter, followed by feature extraction (maximum, mean, and area under the curve). Feature selection was performed using the Mutual Information method. The data were then classified into two categories (fresh and used) using four machine learning algorithms: Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Gradient Boosting, and Random Forest. Model evaluation was carried out using Leave-One-Out Cross Validation (LOOCV) and confusion matrix analysis. The results showed that the Random Forest model achieved the highest testing accuracy of up to 100%, with stable performance even when the number of features was reduced. These findings demonstrate that the combination of E-Nose and machine learning provides an effective, rapid, and non-destructive method for distinguishing between fresh and used cooking oil.

Item Type:	Thesis (Sarjana)
Additional Information:	1). Dr.rer.nat. Bambang Heru Iswanto, M.Si. ; 2). Fachriza Fathan, M.Si.
Subjects:	Sains > Fisika
Divisions:	FMIPA > S1 Fisika
Depositing User:	Chairunisa Nurul Sadiah
Date Deposited:	27 Aug 2025 03:03
Last Modified:	27 Aug 2025 03:03
URI:	http://repository.unj.ac.id/id/eprint/62127

Actions (login required)

View Item