Comparative Study of Deep Learning Methods LSTM and 1D CNN Algorithm: Case Study Of Air Pollution Standard Index Data in DKI Jakarta

POMALINGO, JODIKAL (2022) Comparative Study of Deep Learning Methods LSTM and 1D CNN Algorithm: Case Study Of Air Pollution Standard Index Data in DKI Jakarta. S1 thesis, Universitas Mercu Buana Jakarta.

[img]
Preview
Text (HAL COVER)
01 Cover.pdf

Download (1MB) | Preview
[img]
Preview
Text (ABSTRAK)
02 Abstrak.pdf

Download (145kB) | Preview
[img] Text (BAB I)
03 Bab 1.pdf
Restricted to Registered users only

Download (231kB)
[img] Text (BAB II)
04 Bab 2.pdf
Restricted to Registered users only

Download (201kB)
[img] Text (BAB III)
05 Bab 3.pdf
Restricted to Registered users only

Download (239kB)
[img] Text (BAB IV)
06 Bab 4.pdf
Restricted to Registered users only

Download (216kB)
[img] Text (BAB V)
07 Bab 5.pdf
Restricted to Registered users only

Download (571kB)
[img] Text (BAB VI)
08 Bab 6.pdf
Restricted to Registered users only

Download (261kB)
[img] Text (DAFTAR PUSTAKA)
09 Daftar Pustaka.pdf
Restricted to Registered users only

Download (180kB)
[img] Text (LAMPIRAN)
10 Lampiran.pdf
Restricted to Registered users only

Download (658kB)

Abstract

Many big cities such as DKI Jakarta are experiencing a clean air crisis due to the development of industrialization, the increase in the number of private cars, and the burning of fossil fuels; Air quality is declining, with air pollution getting worse. The problem of handling air pollution is still the government's main focus because the impact of air pollution affects daily life, causing severe health problems for humans and other living things. We need a system that can classify air quality based on air pollution standard index parameters to become a consideration for the government in making air pollution control decisions. Therefore, this study aims to classify air quality based on parameters that affect air quality based on air quality categories. In addition, it also compares the Long Short-Term Memory (LSTM) and One Dimensional Convolutional Neural Network (1D CNN) algorithms. Based on the experiments that have been carried out, the LSTM algorithm outperforms CNN 1D. The results show that both algorithms provide significant accuracy results equally well. The cross-validation results show that the LSTM algorithm obtains the best accuracy of 98.67% on a 5-fold cross-validation with an execution time of 315,049s. At the same time, the CNN 1D algorithm obtained the best accuracy of 98.08% with a time of 416.74s. LSTM provides better accuracy values for fewer kfolds. In comparison, CNN 1D obtained a better accuracy value at a larger k-fold. In conclusion, with the type of quantitative data and the characteristics of the low level of data variation, LSTM can be the proper method in classifying air quality. Key words: Deep Learning, LSTM, 1D CNN, classification, ISPU Banyak kota besar seperti DKI Jakarta mengalami krisis udara bersih akibat perkembangan industrialisasi, peningkatan jumlah mobil pribadi, dan pembakaran bahan bakar fosil; Kualitas udara menurun, dengan polusi udara yang semakin buruk. Masalah penanganan pencemaran udara masih menjadi fokus utama pemerintah karena dampak pencemaran udara mempengaruhi kehidupan seharihari sehingga menimbulkan gangguan kesehatan yang parah bagi manusia dan makhluk hidup lainnya. Diperlukan suatu sistem yang dapat mengklasifikasikan kualitas udara berdasarkan parameter indeks standar pencemaran udara untuk menjadi pertimbangan bagi pemerintah dalam mengambil keputusan pengendalian pencemaran udara. Oleh karena itu, penelitian ini bertujuan untuk mengklasifikasikan kualitas udara berdasarkan parameter yang mempengaruhi kualitas udara berdasarkan kategori kualitas udara. Selain itu, juga membandingkan algoritma Long Short Term Memory (LSTM) dan One Dimensional Convolutional Neural Network (1D CNN). Berdasarkan percobaan yang telah dilakukan, algoritma LSTM mengungguli CNN 1D. Hasil penelitian menunjukkan bahwa kedua algoritma memberikan hasil akurasi yang signifikan sama baiknya. Hasil cross-validation menunjukkan bahwa algoritma LSTM memperoleh akurasi terbaik sebesar 98,67% pada cross-validation 5 kali lipat dengan waktu eksekusi 315,049s. Pada saat yang sama, algoritma CNN 1D memperoleh akurasi terbaik sebesar 98,08% dengan waktu 416,74s. LSTM memberikan nilai akurasi yang lebih baik untuk k-fold yang lebih sedikit. Sebagai perbandingan, CNN 1D memperoleh nilai akurasi yang lebih baik pada k-fold yang lebih besar. Kesimpulannya, dengan jenis data kuantitatif dan karakteristik tingkat variasi data yang rendah, LSTM dapat menjadi metode yang tepat dalam mengklasifikasikan kualitas udara. Kata kunci: Deep Learning, LSTM, 1D CNN, classification, ISPU

Item Type: Thesis (S1)
Call Number CD: FIK/INFO. 22 012
NIM/NIDN Creators: 41518010033
Uncontrolled Keywords: Deep Learning, LSTM, 1D CNN, classification, ISPU
Subjects: 000 Computer Science, Information and General Works/Ilmu Komputer, Informasi, dan Karya Umum > 000. Computer Science, Information and General Works/Ilmu Komputer, Informasi, dan Karya Umum
000 Computer Science, Information and General Works/Ilmu Komputer, Informasi, dan Karya Umum > 000. Computer Science, Information and General Works/Ilmu Komputer, Informasi, dan Karya Umum > 004 Data Processing, Computer Science/Pemrosesan Data, Ilmu Komputer, Teknik Informatika
600 Technology/Teknologi > 600. Technology/Teknologi
Divisions: Fakultas Ilmu Komputer > Informatika
Depositing User: Dede Muksin Lubis
Date Deposited: 08 Jul 2022 07:20
Last Modified: 08 Jul 2022 07:20
URI: http://repository.mercubuana.ac.id/id/eprint/64824

Actions (login required)

View Item View Item