To view the notebook in nbviewer click here
Project completed as part of the Imperial College Business School MSc Business Analytics course. Developing a custom centrality measure to identify important members of a social network and their influence over others in the network.
A high-school Suicide and Distress Counseling service has data on the phone calls between 27 high-school students (a who-talks-to-whom network). One of its directors believes network analytics will help it proactively monitor and help students in distress.
One of the concerns in sociology of young adults is the prevalence of suicide-pacts between pairs of friends. The Director wants to come up with a measure so that if they identify a person as potentially troubled, they can identify a “paired” subject that they can reach out pro-actively to assess suicide risk. It is known that potential at-risk students become increasingly withdrawn
Given this information, develop a centrality measure which identifies the most at risk students in the network.
*unfortunately the raw data used for this project cannot be included in this repository.
Libraries used
-
Networkx
-
Pandas
-
Numpy
-
Matplotlib
-
Seaborn