Studer, Baptiste (2016) Causal Network Inference in Time Series PRE - Research Project, ENSTA.



In recent years, there has been a growing need for inference of underlying causal network in dynamic processes, as time series are more and more used and because the computational power of computers have greatly increased. Many fields of research such as biology, sociology and engineering are involved in this major concern. The purpose of infering underlying causal network is to understand the functionning of dynamic processes and ultimately optimize their performance. This research project focuses on the issue faced by the city of Belfast with its road network. Measurement of traffic flow have been performed on the traffic over few days to study the car distribution between the different roads of the network. The work done was to implement an incremental algorithm which is able to infer the traffic distribution at the different intersections of the road network using the flow measurement carried out.

Item Type:Thesis (PRE - Research Project)
Uncontrolled Keywords:Causal inference, time series, Granger causality
Subjects:Mathematics and Applications
ID Code:6706
Deposited By:Baptiste Studer
Deposited On:06 sept. 2016 15:28
Dernière modification:06 sept. 2016 15:30

Repository Staff Only: item control page