Consider the graph assigned to you. 5 5 5 5 Devin Johnson 2 6 Bobby Sanders...

Transcribed Image Text:

Consider the graph assigned to you. 5 5 5 5 Devin Johnson 2 6 Bobby Sanders 6 3 Demetric Baines 6 Paris Smith 8 8 8 5 5 5 6 Emmanuel Onung 6 Jaden Stanton 3 6 James Ejiwale 8 8 8 5 Julian Parker 5 5 3 6 Cameron Weems 2 6 Nikitha Ketineni 8 Q3: 15 pts) Determine the Spearman's rank-based correlation coefficient between BWC and LCC'DC based on the results of Q2. Show all the work. 4 Q1: 25 pts) Determine the LCC'DC of the vertices in the graph (without using the Jar files) on the basis of the egocentric networks of the vertices. Note that the LCC of a vertex with just one neighbor is 1.0. Q4: 15 pts) Determine the Kendall's concordance-based correlation coefficient between BWC and LCC'DC. Show all the work. 8 Q2: 5 pts) Use the CentralityMetrics.jar file to input the edge list of the graph and obtain the values with respect to the five centrality metrics (DEV, EVC, BWC, CLC and LCC'DC). Q5: 15 pts) Determine the Pearson's correlation coefficient between BWC and LCC'DC. Show all work. Q6: 25 pts) Fit a line (using Excel) that relates LCC'DC and BWC. This is linear regression. Use the equation of the fitted line and the actual LCC'DC values to predict the BWC values of the vertices. Use the predicted BWC and the actual BWC values to determine the standard error residuals (SER). Consider the graph assigned to you. 5 5 5 5 Devin Johnson 2 6 Bobby Sanders 6 3 Demetric Baines 6 Paris Smith 8 8 8 5 5 5 6 Emmanuel Onung 6 Jaden Stanton 3 6 James Ejiwale 8 8 8 5 Julian Parker 5 5 3 6 Cameron Weems 2 6 Nikitha Ketineni 8 Q3: 15 pts) Determine the Spearman's rank-based correlation coefficient between BWC and LCC'DC based on the results of Q2. Show all the work. 4 Q1: 25 pts) Determine the LCC'DC of the vertices in the graph (without using the Jar files) on the basis of the egocentric networks of the vertices. Note that the LCC of a vertex with just one neighbor is 1.0. Q4: 15 pts) Determine the Kendall's concordance-based correlation coefficient between BWC and LCC'DC. Show all the work. 8 Q2: 5 pts) Use the CentralityMetrics.jar file to input the edge list of the graph and obtain the values with respect to the five centrality metrics (DEV, EVC, BWC, CLC and LCC'DC). Q5: 15 pts) Determine the Pearson's correlation coefficient between BWC and LCC'DC. Show all work. Q6: 25 pts) Fit a line (using Excel) that relates LCC'DC and BWC. This is linear regression. Use the equation of the fitted line and the actual LCC'DC values to predict the BWC values of the vertices. Use the predicted BWC and the actual BWC values to determine the standard error residuals (SER).

Answer rating: 100% (QA)

LCCDC stands for Local Clustering Coefficient Degree Correlation Its a metric used in social network analysis to measure a vertexs importance based on its local neighborhood Heres how to calculate LCCDC for each vertex without using JAR files Identify the degree of each vertex The degree is the number of connections a vertex has to other vertices In the image Devin Johnson has a degree of 4 since he has four connections Identify the local clustering coefficient LCC of each vertex LCC is a measure of how wellconnected a vertexs neighbors are to each other A vertex with a high LCC has neighbors that are also connected to each other Calculating LCC can be a bit more involved Heres a general outline Identify the number of edges between the vertexs neighbors Divide the number of edges between neighbors by the total number of possible edges between neighbors For a vertex with n neighbors the total number of possible edges between neighbors is nn12 Once you have the degree and LCC for each vertex multiply the degree by 1 LCC Thats the LCCDC for that vertex For example lets say Devin Johnson vertex 1 has an LCC of 05 His LCCDC would be 4 1 05 2 Note Calculating LCC can be tedious especially for vertices with many neighbors Deeper Dive into LCCDC Calculation Heres a more detailed explanation of calculating LCCDC for each vertex in a graph 1 Degree Calculation This is fairly straightforward For each vertex simply count the number of edges connections it has to other vertices In your social network example you can visually count the edges connected to each person 2 Local Clustering Coefficient LCC Calculation This involves analyzing the connections between a vertexs neighbors Heres a stepbystep breakdown Identify a Vertex Pick a vertex you want to calculate LCC ... View the full answer