Prediksi Konsumsi Gas Alam Rumah Tangga Di Kota Surabaya Menggunakan Model Decision Tree

Triani, Yanuar (2022) Prediksi Konsumsi Gas Alam Rumah Tangga Di Kota Surabaya Menggunakan Model Decision Tree. Masters thesis, Institut Teknologi Sepuluh Nopember.

	Text 09211950094011-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2024. Download (2MB) | Request a copy
	Text 09211950094011-Master_Thesis.pdf Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Gas alam merupakan salah satu sumber yang dihasilkan, baik dari sumber energi terbarukan maupun sumber energi tidak terbarukan. Namun pemanfaatan gas alam yang belum maksimal, khususnya untuk rumah tangga, membuat pemerintah mempercepat perluasan sistem jaringan gas. Penelitian ini bertujuan untuk merumuskan model prediksi yang akurat terkait konsumsi gas alam rumah tangga. PT Perusahaan Gas Negara Tbk. (“PGN”) memerlukan strategi sebagai dasar dalam merencanakan pengembangan di Kota Surabaya, untuk meningkatkan konsumsi gas alam di kelompok rumah tangga. Metode yang digunakan adalah Decision Tree untuk memahami faktor-faktor konsumsi gas alam rumah tangga sekaligus menguji akurasi model menggunakan metode MAE, MAPE, dan RMSE. Hasil penelitian menunjukkan bahwa penggunaan metode Decision Tree terbukti akurat yang ditunjukkan dengan nilai MASE, RMSE, MAE yang menunjukkan sedikit perbedaan hasil dalam memperoleh hubungan antara faktor karakteristik bangunan, faktor sosial demografi, dan faktor psikologi. PGN di Kota Surabaya dalam menentukan kelompok sasaran yang tepat untuk meningkatkan konsumsi gas alam terutama pada rumah tangga dapat menyasar kelompok dengan memperhatikan tingkat pendapatan, tipe pelanggan, usia, status pekerjaan ibu/bapak rumah tangga, pegawai negeri, pegawai swasta, pelanggan yang sangat peduli mengenai harga dan lainnya, serta kelompok yang mengutamakan nilai-nilai CER, altruistik, biosfer, norma sosial, egoistik dan identitas diri pro lingkungan.
================================================================================================
Natural gas is one of the sources produced, both from renewable energy sources and non-renewable energy sources. However, the utilization of natural gas that has not been maximized, especially for households, has made the government accelerate the expansion of the gas network system. This study aims to formulate an accurate predictive model related to household natural gas consumption. PT Perusahaan Gas Negara Tbk. (“PGN”) requires a strategy as a basis for planning development in Surabaya to increase the consumption of natural gas in the household group. The method used is a Decision Tree to understand the factors of household natural gas consumption and test the accuracy of the model using the MAE, MAPE, and RMSE methods. The results showed that the Decision Tree method was proven to be accurate as indicated by the MASE, RMSE, MAE values, which showed a slight difference in obtaining the relationship between building characteristics, socio-demographic, and psychological factors. PGN in Surabaya, especially in the household segmentation, in order to determine the right target group to increase natural gas consumption, can target groups by considering, such as income level, type of customer, age, employment status of housewife, civil servant, a private employee and customers who care about prices and others and who prioritize CER values, altruistic, biosphere, social norms, egoistic, and pro-environmental self-identity.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Decision Tree, Household Consumption, MAE, MAPE, Natural Gas, RMSE, Decision Tree, Gas Alam, Konsumsi Rumah Tangga
Subjects:	H Social Sciences > HA Statistics > HA31.3 Regression. Correlation
Divisions:	Faculty of Creative Design and Digital Business (CREABIZ) > Technology Management > 61101-(S2) Master Thesis
Depositing User:	Yanuar Triani
Date Deposited:	14 Feb 2022 01:06
Last Modified:	02 Nov 2022 04:14
URI:	http://repository.its.ac.id/id/eprint/93809

Actions (login required)

View Item