eprintid: 45495 rev_number: 22 eprint_status: archive userid: 19331 dir: disk0/00/04/54/95 datestamp: 2025-01-17 09:38:26 lastmod: 2025-01-17 09:38:26 status_changed: 2025-01-17 09:38:26 type: thesis metadata_visibility: show contact_email: 3332200060@untirta.ac.id creators_name: SOFYAN ABIYYU, MUHAMMAD creators_id: 3332200060 contributors_type: http://www.loc.gov/loc.terms/relators/THS contributors_type: http://www.loc.gov/loc.terms/relators/THS contributors_name: Ahendyarti, Ceri contributors_name: Muhammad, Fadil contributors_id: 199003042019032012 contributors_id: 199104172019031013 corp_creators: UNIVERSITAS SULTAN AGENG TIRTAYASA corp_creators: FAKULTAS TEKNIK corp_creators: JURUSAN TEKNIK ELEKTRO title: ANALISIS PERFORMA MACHINE LEARNING BERBASIS CONVOLUTIONAL NEURAL NETWORK PADA APLIKASI ECOTRACK ispublished: pub subjects: TK subjects: TL subjects: TR divisions: FT divisions: Elektro full_text_status: restricted keywords: Klasifikasi kendaraan bermotor dan prediksi emisi gas CO2 adalah dua masalah penting dalam upaya mengurangi dampak lingkungan dari sektor transportasi. Penelitian ini bertujuan untuk mengembangkan model berbasis machine learning yang menggunakan arsitektur Convolutional Neural Network (CNN), Artificial Neural Network (ANN), dan transfer learning untuk melakukan klasifikasi kendaraan mobil dan memprediksi emisi gas CO2 yang dihasilkan. Model CNN dan transfer learning yang diusulkan dilatih dengan data gambar kendaraan untuk mengklasifikasikan jenis kendaraan berdasarkan kategori seperti ukuran, jenis, dan tipe kendaraan. Hasil klasifikasi kendaraan memiliki tingkat akurasi yang tinggi mencapai 85,14% dan 68,02% pada model model transfer learning yang dibuat. Hasil klasifikasi kendaraan memiliki tingkat akurasi yang rendah pada model CNN yaitu sebesar 43%. Hal ini diakibatkan tidak optimalnya algoritma machine learning. Selain itu, model ini juga dilatih menggunakan data emisi kendaraan untuk memprediksi emisi CO2 yang dihasilkan berdasarkan parameter kendaraan seperti kapasitas mesin, silinder mesin, kombinasi konsumsi bahan bakar, dan estimasi jumlah emisi gas CO2 yang dihasilkan kendaraan bermotor. Hasil penelitian menunjukkan bahwa model CNN dan transfer learning mampu dalam klasifikasi kendaraan serta model ANN memberikan estimasi yang akurat untuk prediksi emisi CO2. abstract: Classification of motor vehicles and prediction of CO2 gas emissions are two important issues in efforts to mitigate the environmental impact of the transportation sector. This study aims to develop a machine learning-based model utilizing Convolutional Neural Network (CNN), Artificial Neural Network (ANN), and transfer learning architectures to classify vehicles and predict the CO2 gas emissions they produce. The proposed CNN and transfer learning models were trained using vehicle image data to classify vehicle types based on categories such as size, type, and model. The results of vehicle classification showed high accuracy rates of 85.14% and 68.02% in the developed transfer learning models. However, the CNN model demonstrated lower accuracy, achieving only 43%, which was attributed to the suboptimal performance of the machine learning algorithm. Furthermore, the model was also trained with vehicle emission data to predict CO2 emissions based on vehicle parameters such as engine capacity, number of cylinders, combined fuel consumption, and the estimated amount of CO2 gas emissions produced by motor vehicles.The findings indicate that the CNN and transfer learning models are effective for vehicle classification, while the ANN model provides accurate estimates for CO2 emission predictions. date: 2024-01-16 date_type: published pages: 102 institution: Fakultas Teknik Universitas Sultan Ageng Tirtayasa department: TEKNIK ELEKTRO thesis_type: sarjana thesis_name: sarjana citation: SOFYAN ABIYYU, MUHAMMAD (2024) ANALISIS PERFORMA MACHINE LEARNING BERBASIS CONVOLUTIONAL NEURAL NETWORK PADA APLIKASI ECOTRACK. S1 thesis, Fakultas Teknik Universitas Sultan Ageng Tirtayasa. document_url: https://eprints.untirta.ac.id/45495/1/Muhammad%20Sofyan%20Abiyyu_3332200060_Fulltext.pdf document_url: https://eprints.untirta.ac.id/45495/2/Muhammad%20Sofyan%20Abiyyu_3332200060_01.pdf document_url: https://eprints.untirta.ac.id/45495/4/Muhammad%20Sofyan%20Abiyyu_3332200060_02.pdf document_url: https://eprints.untirta.ac.id/45495/3/Muhammad%20Sofyan%20Abiyyu_3332200060_03.pdf document_url: https://eprints.untirta.ac.id/45495/5/Muhammad%20Sofyan%20Abiyyu_3332200060_04.pdf document_url: https://eprints.untirta.ac.id/45495/6/Muhammad%20Sofyan%20Abiyyu_3332200060_05.pdf document_url: https://eprints.untirta.ac.id/45495/7/Muhammad%20Sofyan%20Abiyyu_33332200060_Ref.pdf document_url: https://eprints.untirta.ac.id/45495/8/Muhammad%20Sofyan%20Abiyyu_3332200060_Lamp.pdf document_url: https://eprints.untirta.ac.id/45495/9/Muhammad%20Sofyan%20Abiyyu_3332200060_CP.pdf