Modeling of Residents' Travel Mode Choice Decisions during Peak Commuting Hours via RL-SVM Method

In response to worsening traffic congestion, cities worldwide have prioritized the development of public transportation systems, striving become 'public transport cities.' To address this issue, conducting an in-depth study residents' travel choice behavior during peak hours is vital importance. The paper aims consider key factors affecting decisions by utilizing revealed preference (RP) samples and proposes a novel RL-SVM model based on random parameter logit (RL) support vector machine (SVM) theories for mode prediction. Comparative experimental analysis shows that our outperforms traditional prediction methods regarding classification sensitivity. Therefore, we can further evaluate anticipated impact implementing priority strategies obtain internal shift patterns commuters' modes from private cars transport. It holds significant implications promoting healthy sustainable urban development.