Kaijie “Yuri” Yu, Ahmad Alasmari, Edward Segura-Tinoco, Denise Escarcega, Ghurdan Alghureid

Data Cleaning: preview.ipynb

SNA: Group_Project.rmd

Presentation: Presentation.pptx

Report: TNDY336_GroupM.docx

In our study, we aim to analyze the HIV transmission network to understand its structure, identify high-risk subgroups, and inform potential interventions. Our research will delve into various aspects, such as centrality measures, community detection, network models, predictive modeling, and identification of key individuals or groups, to provide a comprehensive understanding of the network and its implications on HIV transmission.

(Weighted out-degree)

1. What is the five-number summary of the network? How can we interpret them?
2. How do centrality measures (degree, betweenness, eigenvector) differ across node attributes? Are certain subgroups more central in the network?
3. Can community detection algorithms identify subgroups within the network based on node attributes, and if there are subgroups with a higher risk of HIV transmission? What are the defining characteristics of these subgroups?
4. Can network models, such as Exponential Random Graph Models (ERGMs), help explain the formation and structure of the HIV transmission network based on node attributes? Which attributes have the strongest impact on network formation and HIV transmission?
5. Can the relationships between node attributes and network properties (e.g., centrality measures) be used to develop predictive models for HIV transmission risk within the network?
6. Can we identify any key individuals or groups of individuals (e.g., opinion leaders, influencers) within the network who could be targeted for interventions to reduce HIV transmission? We can use algorithms like k-shell decomposition or methods like snowball sampling to identify these individuals or groups.
7. Can we identify any structural holes or brokerage positions within the network, and how do these positions relate to node attributes? Individuals occupying these positions may have a unique role in the information or resource flow within the network. We can use methods like Burt's constraint measure to identify structural holes or brokerage positions.