Don't miss our weekly PhD newsletter | Sign up now Don't miss our weekly PhD newsletter | Sign up now

  Coupling Artificial Intelligence with Statistical and Econometric Methods for Transport Data Analysis

   School of Computing, Engineering & the Built Environment

This project is no longer listed on and may not be available.

Click here to search for PhD studentship opportunities
  Dr G Fountas, Dr A Fonzone  No more applications being accepted  Self-Funded PhD Students Only

About the Project

The emergence of novel data collection technologies and techniques has paved the way for the widespread use of “big data”, which can better inform decisions in transport planning, operations and policy. Real- time data, naturalistic and sensor data, or even social media and app data constitute a few representative examples of such new sources of disaggregate information. The high dimensionality of this data in combination with the extensive range of newly featured variables pose significant challenges in their analysis and modeling.

Over the last few years, the rise of Artificial Intelligence (AI) methods has brought unprecedented capabilities in efficiently handling and processing large datasets, with their use typically resulting in high forecasting accuracy. However, the long-debated “black-box” components of various AI approaches set profound restrictions in their ability to unravel causality effects from the analysis of “big data”, thus leading to low explanatory power and limited application potential.

In contrast, the traditional statistical and econometric methods and their recent extensions have proven more efficient in identifying underlying relationships between data elements, allowing also for capturing observed or unobserved effects. However, the complexities underpinning their estimation process impose significant barriers to their convergence, especially in cases of large and disparate datasets.

This PhD project seeks to develop a hybrid data analysis framework, which will integrate co-benefits of AI and statistical and econometric methods for the alignment and modeling of disparate transport data. Such a framework has the potential to provide a decisive step towards resolving the main dilemma transport analysts and researchers face in the selection process of the most appropriate data analysis approach:

“What should my model do? Predict or explain”? This PhD program is anticipated to foster the coupling of AI and statistical econometric methods in order to jointly optimise the predictive and explanatory power of data-driven models in transport analyses. In this context, this research will contribute to answering the previous question with a resounding “Both”.

The successful candidate will join the Transport Research Institute (TRI) of Edinburgh Napier University, the Scotland’s largest and longest established transport research group.

Academic qualifications

A first degree (at least a 2.1) ideally in statistics/econometrics/applied mathematics/data science/computer science/transportation or industrial engineering with a good fundamental knowledge of statistical and econometric methods and/or artificial intelligence techniques for transport data analysis.

English language requirement

IELTS score must be at least 6.5 (with not less than 6.0 in each of the four components). Other, equivalent qualifications will be accepted. Full details of the University’s policy are available online.

Essential attributes: · Experience of fundamental knowledge of statistical and econometric methods and/or artificial intelligence in data analysis. · Competent in computer programming and advanced data analysis. · Knowledge of transport analysis and modelling, especially in areas related to travel behaviour and/or traffic safety. · Good written and oral communication skills · Strong motivation, with evidence of independent research skills relevant to the project · Good time management Desirable attributes: Previous experience in statistical modelling of transport data Previous experience in using Artificial Intelligence methods for transport analysis Familiarity with big data analytics

Engineering (12) Mathematics (25)


Mannering, F., Bhat, C.R., Shankar, V., Abdel-Aty, M., 2020. Big data, traditional data and the tradeoffs between prediction and causality in highway-safety analysis. Analytic methods in accident research, 25, 100113. Karlaftis, M.G. and Vlahogianni, E.I., 2011. Statistical methods versus neural networks in transportation research: Differences, similarities and some insights. Transportation Research Part C: Emerging Technologies, 19(3), pp.387-399. Brathwaite, T., 2018. The Holy Trinity: Blending Statistics, Machine Learning and Discrete Choice, with Applications to Strategic Bicycle Planning (Doctoral dissertation, UC Berkeley). Washington, S., Karlaftis, M.G., Mannering, F. Anastasopoulos, P., 2020. Statistical and econometric methods for transportation data analysis. CRC press. Fountas, G., Anastasopoulos, P.C., Boyle, L., 2019. Opportunities and Challenges in Statistical Analysis of Transportation Data: Where We Are and Where We Are Going. TRB Centennial Papers.

 About the Project