Introduction:
Everyday, I, along with millions of other, take the subway on my way to work and school. The total global number of people whose use the subway is expected to continue to increase as the world recovers from Covid and the overall world population continues to increase. Because so many people are depend on quick and reliable metro systems, it is important to make them as efficient and as durable as possible.
Analysis:
In this study, researchers looked at metro networks in Beijing, Hong Kong, London, Paris and Tokyo (imagine a graph where each station is a vertex and each track between the stations as an edge) and analysed each them to judge their reliable and efficiency. In this blog, I’ll focus on the two metrics that relate the most to the course material: The clustering coefficient and network efficiency
Average Clustering Coefficient
One metric the researchers looked at was the average clustering coefficient of each network. The average clustering coefficient is calculated by adding the clustering coefficient of each station and dividing it by the total number of stations. Having a high clustering coefficient is important because it makes the network more resistant to track / station closures since the neighbors of the closed station are more likely to be connected to each other meaning commuters can find alternate roots that bypass any closed station.
Global Network Efficiency
Another metric the researcher’s analysed was the global network efficiency which is the inverse of the average shortest path commute between any two stations in the metro network. The higher a metro’s network efficiency is, the quicker the average commute is expected to be.
Results
The study suggests that the London subway was the most reliable metro network due to it’s high average clustering coefficient while the Tokyo subway had the highest global network efficiency.
Source:
Wu, X., Dong, H., Ho, I., Lau, F., & Tse, C. (2016, November 30). A network analysis of World’s Metro Systems – Institute of Electronics … ieice. Retrieved October 7, 2022, from https://www.ieice.org/nolta/symposium/archive/2016/articles/1065.pdf