Implementasi Traffic Forecasting dengan Machine Learning dan Explainable Artificial Intelligence pada Aplikasi berbasis Website

Pramono, Damarhafni Rahmannabel Nadim (2024) Implementasi Traffic Forecasting dengan Machine Learning dan Explainable Artificial Intelligence pada Aplikasi berbasis Website. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5027201026-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 July 2026. Download (5MB) | Request a copy

Abstract

Menurut Organisasi Kesehatan Dunia (WHO), kecelakaan lalu lintas adalah kejadian tak terduga yang menyebabkan cedera dan telah menjadi perhatian global, baik secara medis maupun ekonomi. Dengan kecelakaan lalu lintas jalan raya yang menjadi penyebab kematian ketujuh dan kesepuluh di seluruh dunia, pentingnya manajemen lalu lintas yang lebih baik telah disoroti. Intelligent Transportation System (ITS) telah dikembangkan untuk mengatasi masalah ini, dengan memanfaatkan analisis Time-Series untuk menghasilkan proyeksi lalu lintas secara real-time. Meskipun demikian, meskipun terdapat terobosan ilmiah yang signifikan dalam proyeksi tersebut, dampaknya akan terbatas jika konsumen tidak memahami dan menerimanya. Studi ini menggabungkan model Machine Learning berbasis ARIMA - metode yang terkenal dalam time series forecasting - dengan Explainable Artificial Intelligence (XAI), dengan tujuan untuk memberikan transparansi dan penjelasan kepada konsumen tentang forecasting lalu lintas. Penelitian ini menggunakan dataset yang tersedia untuk umum dari tahun 2012 hingga 2017, serta persiapan dataset, Explanatory Data Analysis (EDA), preprocessing data untuk training dan testing, dan proses pemodelan ARIMA. Setelah prosedur pemodelan, XAI digunakan untuk meningkatkan transparansi sebelum mengembangkan aplikasi berfokus pada pengguna yang menggambarkan hasil forecasting. Hasil penelitian ini menunjukkan bahwa model terbaik merupakan model XGBoost dengan nilai metrik evaluasi RMSE 292,0429, MSE 85.289,0438, MAE 190,2757, dan MAPE 8,4087. Model yang diteliti juga dapat menggunakan XAI berupa SHAP dan LIME untuk menjelaskan transparansi model.
====================================================================================================================================
According to the World Health Organization (WHO), traffic accidents are unexpected events that cause injury and have become a global concern, both medically and economically. With road traffic accidents becoming the seventh and tenth top cause of mortality worldwide, the significance of improved traffic management has been emphasized. The Intelligent Transportation System (ITS) was launched in this respect, leveraging Time-Series analysis to produce real-time traffic projections. Nonetheless, despite significant scientific breakthroughs in such projections, their impact will be limited if consumers do not comprehend and accept them. This study combines ARIMA-based Machine Learning models - a well-known method in time-series forecasting - with Explainable Artificial Intelligence (XAI), with the goal of providing consumers with transparency and explanation of traffic predictions. This study makes use of publicly available datasets from 2012 to 2017, as well as dataset preparation, Explanatory Data Analysis (EDA), data preprocessing for training and testing, and ARIMA modeling processes. Following the modeling procedure, XAI was used to increase transparency before developing a user-focused application depicting the forecasting results. The results of this study show that the best model is the XGBoost model with an evaluation metric value of RMSE 292.0429, MSE 85,289.0438, MAE 190.2757, and MAPE 8.4087. The model studied can also use XAI in the form of SHAP and LIME to explain the transparency of the model.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Analisis Time-Series, ARIMA, Explainable Artificial Intelligence (XAI), Intelligent Transportation System (ITS), Kecelakaan Lalu Lintas, Machine Learning, Traffic Forecasting; Time-Series Analysis, ARIMA, Explainable Artificial Intelligence (XAI), Intelligent Transportation System (ITS), Traffic Accident, Machine Learning, Traffic Forecasting.
Subjects:	T Technology > T Technology (General) > T174 Technological forecasting T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105.888 Web sites--Design. Web site development.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information Technology > 59201-(S1) Undergraduate Thesis
Depositing User:	Damarhafni Rahmannabel Nadim Pramono
Date Deposited:	02 Feb 2024 06:31
Last Modified:	02 Feb 2024 06:31
URI:	http://repository.its.ac.id/id/eprint/105961

Actions (login required)

View Item