Analisa Efektivitas Model Machine Learning Dalam Prediksi Kandungan Formalin Pada Mie Basah

Lodianto, Bagas Immanuel (2022) Analisa Efektivitas Model Machine Learning Dalam Prediksi Kandungan Formalin Pada Mie Basah. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 05311840000026-Undergraduate_Thesis.pdf - Accepted Version Download (2MB)

Abstract

Mie merupakan makanan yang sering dikonsumsi sebagai alternatif nasi sebagai sumber karbohidrat. Akan tetapi mie memiliki masa simpan makanan yang tidak tahan lama dan dengan mudahnya akses beli berbagai macam pengawet yang berada di pasar, tidak sedikit pedagang menyalahgunakan zat-zat berbahaya seperti formalin sebagai pengawet. Telah banyak metode untuk mendeteksi formalin di dalam suatu makanan, akan tetapi banyak dari proses tersebut membutuhkan waktu dan biaya yang cukup besar. Pengembangan suatu alat bantu untuk memudahkan deteksi formalin dalam mie basah secara mudah dan cepat dapat menjadi solusi untuk permasalahan ini. Penggunaan sensor gas bersama dengan machine learning merupakan kunci utama dari proses prediksi kandungan formlain dalam mie basah. Electronic nose atau biasa disebut dengan E-Nose dibangun menggunakan Arduino UNO dan beberapa gas sensor MQ. Untuk mendapatkan hasil terbaik dalam prediksi kandungan formalin, maka dilakukan perbandingan terhadap tiga macam model yaitu, linear regression, regression trees, dan neural network. Model prediksi dengan performa terbaik yang didapatkan adalah regression trees dengan nilai RMSE dan MAE yang yaitu 5.34 and 3.69. Hasil yang cukup memuaskan ini menunjukkan bahwa E-Nose dapat digunakan sebagai alat bantu alternatif dalam proses prediksi kandungan formalin dalam mie basah.
====================================================================================================================================
Wet noodles are foods that are usually consumed as an alternative for rice to fulfill carbohydrate needs. However, Noodles lifespan is considered short, and with the ease of access to purchase various preservative such within the market, some merchants knowingly misappropriate the use of formalin within foods. There have been many studies that shows the danger of formalin consumption. One of the use cases of this substance is to preserve a wet noodle by soaking it. As of today, there has been many ways to detect formalin content within a food. However, those methods often times require a lot of time and is not considered cheap. The development of a tool that could help detect formalin content within wet noodles that are easy and fast can be a solution for this particular case. The use of MOS gas sensor and machine learning are the main key to predict formalin content within wet noodles. Electronic Nose or in short E-Nose, will consist of Arduino UNO and multiple MOS gas sensor, namely MQ sensor. To achieve best performance to predict formalin content within wet noodles, Three different models will be compared to each other, those models are namely linear regression, regression trees, and neural network. The best predictive model performance was obtained by regression trees with 5.34 and 3.69 for RMSE and MAE score. These satisfactory results demonstrated that E-nose could be applied as an alternative to predict formalin content within wet noodles.

Item Type:	Thesis (Other)
Additional Information:	RSTI 006.312 Lod a-1 2022
Uncontrolled Keywords:	formalin, regresi, MOS sensor, machine learning, formalin, regression,MOS sensor,machine learning
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information Technology > 59201-(S1) Undergraduate Thesis
Depositing User:	- Davi Wah
Date Deposited:	08 Jul 2024 05:15
Last Modified:	08 Jul 2024 05:15
URI:	http://repository.its.ac.id/id/eprint/108004

Actions (login required)

View Item