{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

# hw4 - (b Compare the algorithms using enron2 i plot...

This preview shows page 1. Sign up to view the full content.

CSE 5800 Mining/Learning and the Internet—HW4 Due Nov 16, Wed, 6:30pm Submit Server: course=cse5800 , project=hw4 1. Implement BridgeCut with four versions: (a) edge with the highest Bridging Centrality ( C Br ( e ) in the paper) (b) vertex with the highest Bridging Centrality ( C Br ( v )) (c) edge with the highest Betweenness (Φ( e )) (d) vertex with the highest Betweenness (Φ( v )) 2. Allow this parameter: (a) density threshold (densityThreshold in the paper) 3. Measure performance using: (a) Davies-Bouldin index (b) Silhouette Coefficient (handout) 4. Use three groups of data sets: (a) toy data sets on the course web site (b) real data sets on the course web site (c) your own data set 5. Disscuss in a report (in pdf): (a) Sensitivity analysis of parameters using enron2.txt: i. vary density threshold ii. calculate each performance measurement, iii. plot performance vs. density threshold iv. discuss the value for density threshold that seems to achieve the highest performance.
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: (b) Compare the algorithms using enron2.txt: i. plot performance vs. density threshold for diﬀerent algorithms ii. plot clustering coeﬃcient vs. number of nodes deleted (up to top 10%) for diﬀerent algorithms [Figure 5b in the paper] iii. discuss the relative performance of diﬀerent algorithms 6. Implementation: (a) use one of these programming languages: C, C++, Java, or LISP. (b) input ﬁle: a ﬁle for vertices and edges (c) two modules: i. BridgeCut: input graph; output: • top edge/vertex when it is removed • for each cluster, output vertices in the cluster ii. Evaluate: input vertices and cluster membership; output performance 7. Submission: (a) source code (b) your data set (c) report in pdf (d) README.txt (how to compile and run your program on code.ﬁt.edu)...
View Full Document

{[ snackBarMessage ]}

Ask a homework question - tutors are online