A new survey paper published on arXiv explores the application of Neural Architecture Search (NAS) in traffic prediction. The paper details how NAS can automate the design of deep learning models, such as Graph Convolutional Networks, Recurrent Neural Networks, and Transformers, to better capture the spatial-temporal structure of traffic data. It categorizes NAS methods into gradient-based, evolutionary, and one-shot weight-sharing approaches, while also highlighting challenges like computational scalability and cross-city generalization. AI
IMPACT Automates deep learning model design for traffic prediction, potentially improving efficiency and generalization in intelligent transportation systems.
RANK_REASON The cluster consists of two identical arXiv preprints detailing a survey paper on a specific research topic.
Read on arXiv cs.NE (Neural & Evolutionary) →
- arXiv
- deep learning
- Evolutionary methods for multidisciplinary optimization applied to the design of UAV systems†
- Gradient-based methods
- Graph Convolutional Networks
- intelligent transportation system
- Neural architecture search
- One-shot weight-sharing methods
- Recurrent Neural Networks
- Spatial-temporal foundation models
- Traffic prediction based on map images for autonomous driving
- transformers
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
