Scholarlink

Accurate prediction of the estimated time of arrival (ETA) for buses is crucial for shuttle companies aiming to enhance profitability and minimize costs. This paper proposes leveraging historical bus route data to predict estimated bus arrivals at specific stations, employing K-Nearest Neighbors (KNN) supervised machine learning (ML) techniques. We develop a simple and interpretable solution that abstains from complexity, as bus routes dynamically adapt to passenger demands.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Predictive Model of Arrival Times for Smart Shuttle Buses in Astana, Kazakhstan

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Darya Taratynova,
  • Assel Kassenova,
  • Bissenbay Dauletbayev,
  • Muammar Al-Shedivat,
  • Eugene Pinsky

摘要

Accurate prediction of the estimated time of arrival (ETA) for buses is crucial for shuttle companies aiming to enhance profitability and minimize costs. This paper proposes leveraging historical bus route data to predict estimated bus arrivals at specific stations, employing K-Nearest Neighbors (KNN) supervised machine learning (ML) techniques. We develop a simple and interpretable solution that abstains from complexity, as bus routes dynamically adapt to passenger demands.