English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Real-time demand prediction is a critical input for dynamic bus routing.While many researchers have developed numerous complex methods to predictshort-term transit demand, the applications have been limited to short, stabletime frames and a few stations. How these methods perform in highly dynamicenvironments has not been studied, nor has their performance beensystematically compared. We built an open-source infrastructure with fivecommon methodologies, including econometric and deep learning approaches, andassessed their performance under stable and highly dynamic conditions. We useda time series from smartcard data to predict demand for the following day forthe BRT system in Bogota, Colombia. The dynamic conditions in the time seriesinclude a month-long protest and the COVID-19 pandemic. Both conditionstriggered drastic shifts in demand. The results reveal that most tested modelsperform similarly in stable conditions, with MAAPE varying from 0.08 to 0.12.The benchmark demonstrated that all models performed significantly worse inboth dynamic conditions compared to the stable conditions. In the month-longprotest, the increased MAAPE ranged from 0.14 to 0.24. Similarly, during theCOVID-19 pandemic, the increased MAAPE ranged from 0.12 to 0.82. Notably, inthe COVID-19 pandemic condition, an LSTM model with adaptive training and amulti-output design outperformed other models, adapting faster to disruptions.The prediction error stabilized within approximately 1.5 months, whereas othermodels continued to exhibit higher error rates even a year after the start ofthe pandemic. The aim of this open-source codebase infrastructure is to lowerthe barrier for other researchers to replicate and reproduce models, facilitatea collective effort within the research community to improve the benchmarkingprocess and accelerate the advancement of short-term ridership predictionmodels.

Public Transit Demand Prediction During Highly Dynamic Conditions: A  Meta-Analysis of State-of-the-Art Models and Open-Source Benchmarking  Infrastructure