Hybrid Models for Forecasting Dublin Bus Delay due to Weather Conditions

Solaisamy Alagarsamy Nagarajan, Loheswari (2025) Hybrid Models for Forecasting Dublin Bus Delay due to Weather Conditions. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

Urban transport systems like Dublin’s often experience delays due to weather conditions and peak hour traffic, which affects commuter satisfaction and transport planning. This research proposes a hybrid forecasting model that combines the Facebook Prophet for trend extraction with a Gated Recurrent Unit(GRU) to capture the short term dependencies and external factors. A real-time dataset was extracted via APIs, Containing historical bus delay data and then weather data. Feature engineering techniques like lagged delays, rolling averages and time based indicators were applied to enhance the dataset. Multiple models were evaluated, including GRU, Prophet with external regressors, Random Forest and a GRU+Prophet hybrid model. In the hybrid model where Prophet’s trend prediction were used as additional inputs for GRU. The hybrid model achieved an MAE of 15.98s, RMSE of 22.87s, and R of 0.44, outperforming the GRU(R²=0.2361) and performing comparably to Prophet (R² =0.4222), but Random forest achieved the highest accuracy (R²=0.9963). It lacks interpretability and the model may be over fitted. Although the hybrid model was not the top performer, the model offers a balance between accuracy and interpretability. Its architecture effectively incorporates seasonality and sequential patterns, making it a robust and transparent.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Singh, Jaswinder UNSPECIFIED
Uncontrolled Keywords:	Urban transport forecasting; Bus delay prediction; Hybrid time series model; Gated Recurrent Unit (GRU); Facebook Prophet; Weather based prediction; Machine learning; Feature engineering; Public transit analytics
Subjects:	Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning H Social Sciences > HE Transportation and Communications > Urban Transportation G Geography. Anthropology. Recreation > GE Environmental Sciences > Earth sciences > Atmospheric science > Meteorology > Weather
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	26 Aug 2026 12:12
Last Modified:	26 Aug 2026 12:12
URI:	https://norma.ncirl.ie/id/eprint/9670

Actions (login required)

View Item