Мақола

A Hybrid Statistical and Machine Learning Approach for Time-Series Forecasting of Traffic Accidents in Tashkent

Sokhobiddin AkhatkulovSamarkand State University Named After Sharof Rashidov,Department of Control Theory and Information Security,Samarkand,UzbekistanIslom YalgoshevSamarkand StateUuniversity Named After Sharof Rashidov,Department of Software Engineering,Samarkand,UzbekistanJamoliddin JabbarovSamarkand State University Named After Sharof Rashidov,Department of Control Theory and Information Security,Samarkand,Uzbekistan

2026

ABI

Аннотация

In this study, a hybrid statistical and machine learning models are proposed for forecasting traffic accidents in the city of Tashkent, the capital of Uzbekistan. The hybrid models are based on the following models: Autoregressive Integrated Moving Average (ARIMA), Linear Regression (LR), Simple Exponential Smoothing (SES), and Extreme Gradient Boosting (XGBoost). The data used in this study were collected monthly from 12 districts of Tashkent from 2018 to 2026, and it is available on https://opendata.tashkent.uz/eng (Accessed on Jan. 20, 2026). The model performances are compared with classical models such as SES and ARIMA. The results show that the proposed hybrid models are better performed in all metrics, MAPE, MAE and RMSE. The key features of this time-series dataset are the time and the total traffic accidents across all districts of Tashkent.

Ҳали таржима қилинмаган

Мавзулар

Traffic Prediction and Management Techniques Hydrological Forecasting Using AI Traffic and Road Safety

Идентификаторлар

DOI: 10.1109/smartindustrycon68821.2026.11492864

Иқтибослар ва манбалар

0 та иқтибос13 та фойдаланилган манба

Кўрсаткичлар — AkademScholar