NetReconDist Collabathon: January 17-19
Event
January 10, 2019

First Annual Network Science Institute Winter Collaboration

January 17th-19th, 2019

The Collabathon is a three-day event at the Network Science Institute. The primary goal for this project is to expand the culture of tinkering and collaboration at NetSI, while at the same time producing high impact research. We will be learning from one another constantly during the event, which perfectly highlights the unique nature of NetSI. There are very few institutions on earth that can garner the amount of interdisciplinary and technical knowledge that we have at NetSI, and finding ways to harness this is going to be a key focus as we progress as an institution.  

“Comparing methods for reconstructing networks from time series data by comparing methods for measuring network similarity”

Across many disciplines, we analyze networks that have been reconstructed or inferred from time series data. Network reconstruction can be achieved through a number of techniques, but different algorithms can output different networks, leaving practitioners uncertain about whether their approach is suitable for describing the system in question. As with other network analysis tools, it appears that no single technique is optimal for inferring network structure from time series data due to the quality and amount of time series data being collected, the nature of the system being modeled, and the types of interactions between nodes in the network (causal, correlational, weighted, etc.).

In this Collabathon, systematically and collaboratively implement a number of network similarity (or graph distance) metrics to induce networks from various real and simulated time series datasets. We will then systematically compare the network reconstruction techniques to characterize approaches that are more likely to infer similar network structures from time series data. The purpose was not to provide estimates of the best network reconstruction technique but rather to compare techniques. Our work will provide a useful heuristic for researchers as they select the tools needed for analyzing complex systems.

Read more
Event
NetReconDist Collabathon: January 17-19

First Annual Network Science Institute Winter Collaboration

January 17th-19th, 2019

The Collabathon is a three-day event at the Network Science Institute. The primary goal for this project is to expand the culture of tinkering and collaboration at NetSI, while at the same time producing high impact research. We will be learning from one another constantly during the event, which perfectly highlights the unique nature of NetSI. There are very few institutions on earth that can garner the amount of interdisciplinary and technical knowledge that we have at NetSI, and finding ways to harness this is going to be a key focus as we progress as an institution.  

“Comparing methods for reconstructing networks from time series data by comparing methods for measuring network similarity”

Across many disciplines, we analyze networks that have been reconstructed or inferred from time series data. Network reconstruction can be achieved through a number of techniques, but different algorithms can output different networks, leaving practitioners uncertain about whether their approach is suitable for describing the system in question. As with other network analysis tools, it appears that no single technique is optimal for inferring network structure from time series data due to the quality and amount of time series data being collected, the nature of the system being modeled, and the types of interactions between nodes in the network (causal, correlational, weighted, etc.).

In this Collabathon, systematically and collaboratively implement a number of network similarity (or graph distance) metrics to induce networks from various real and simulated time series datasets. We will then systematically compare the network reconstruction techniques to characterize approaches that are more likely to infer similar network structures from time series data. The purpose was not to provide estimates of the best network reconstruction technique but rather to compare techniques. Our work will provide a useful heuristic for researchers as they select the tools needed for analyzing complex systems.

read more