Statistical methods versus neural networks in transportation research: Differences, similarities and some insights

By Matthew G. Karlaftis, Eleni I. Vlahogianni. 2011

841 citations recorded by the source corpus.

Subfields: Building and Construction, Transportation, Signal Processing

View Paper: https://www.semanticscholar.org/paper/e5fbd231b7263f3e778ba6a939ce68c9a6478e21

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Building and Construction

1. Yinhai Wang
2. Zhiyong Cui
3. Ruimin Ke
4. Weibin Zhang
5. Yunpeng Wang
6. Zhengbing He
7. Xiaolei Ma
8. Shengna Guo

In Transportation

1. Yunpeng Wang
2. Shengna Guo
3. Youfang Lin
4. Huaiyu Wan
5. Xiaolei Ma
6. Yinhai Wang
7. Yinghao Yu
8. Weibin Zhang

In Signal Processing

1. Yinhai Wang
2. Zhiyong Cui
3. Ruimin Ke
4. Eugene Vinitsky
5. Xiang Yan
6. Alexandre M. Bayen
7. Alan Yu
8. Cathy Wu

Citations in the corpus

• Deep Spatial–Temporal 3D Convolutional Neural Networks for Traffic Data Forecasting (2019, background): “Although traditional machine learning methods have solid mathematical foundations and can help us to understand the mechanisms of data generation, they fail to make accurate predictions when dealing with complex and highly nonlinear data [16] and need careful feature engineering.”
• A Survey on Modern Deep Neural Network for Traffic Prediction: Trends, Methods and Challenges (2020, methodology): “In this section, we will first state several of the challenges outlined by Vlahogianni et al. in their 2014 survey paper [9] and discuss only the challenges that have been solved or partially solved using deep neural network.”
• Spatiotemporal Recurrent Convolutional Networks for Traffic Prediction in Transportation Networks (2017, background): “An NN can model complex nonlinear problems with remarkable performance in handling multi-dimensional data [28].”
• Predicting Short-Term Traffic Flow by Long Short-Term Memory Recurrent Neural Network (2015, background): “Artificial Neural Network(ANN) is another popular countermeasure for traffic prediction due to its capability of handling multi-dimensional data, flexible model sturcture, strong generalization and learning ability as well as adaptability [4].”
• Short-term traffic flow prediction based on spatio-temporal analysis and CNN deep learning (2019, background): “Neural networks, as a type of nonparametric model, have the characteristics of high flexi-bilityandgoodlearning,generalization,andpredictionabilities,andhavebeenwidelyused in the field of transportation (Karlaftis and Vlahogianni 2011).”
• Flow: A Modular Learning Framework for Mixed Autonomy Traffic (2017, methodology): “For additional uses of deep learning in traffic, we refer the reader to Karlaftis, et al. [91], which presents an overview comparing non-neural statistical methods and neural networks in transportation research.”
• An Improved Fuzzy Neural Network for Traffic Speed Prediction Considering Periodic Characteristic (2017, background): “approaches are more capable of processing outliers, missing and noisy data [39].”
• Stacked Bidirectional and Unidirectional LSTM Recurrent Neural Network for Forecasting Network-wide Traffic State with Missing Values (2020, methodology): “…the ability to deal with high dimensional data and the capability of capturing non-linear relationship, CI approaches, especially novel machine learning methods, tend to outperform the statistical methods with respect to handling complex traffic forecasting problems Karlaftis and Vlahogianni (2011).”
• Prediction and behavioral analysis of travel mode choice: A comparison of machine learning and logit models (2020, background): “Machine-learning studies rarely apply model outputs to facilitate behavioral interpretation, i.e., to test the response of the output variable or to changes in the input variables in order to generate findings on individual travel behavioral and preferences (Karlaftis and Vlahogianni, 2011).”
• Learning Traffic as Images: A Deep Convolutional Neural Network for Large-Scale Transportation Network Speed Prediction (2017, background): “Artificial neural networks (ANNs) are also usually applied to traffic prediction problems because of its advantages, such as their capability to work with multi-dimensional data, implementation flexibility, generalizability, and strong forecasting power [3].”
• Deep Bidirectional and Unidirectional LSTM Recurrent Neural Network for Network-wide Traffic Speed Prediction (2018, methodology): “With the ability to deal with high dimensional data and the capability of capturing non-linear relationship, CI approaches tend to outperform the statistical methods, such as auto-regressive integrated moving average (ARIMA) [36], with respect to handling complex traffic forecasting problems [38].”