Project Description:
Visualize the netflix dataset in a 2-dimensional manner such that we can glean useful information about the data.

Goals:

• Display the data in a 2-dimensional manner.
• Overcome issues such as lack of adequate coordinate system to display data.
• Implement force-directed graphing algorithm (FDGA).
• Analyze output graphs for useful information.

Sources:

• Fruchterman, Reingold on Force directed placement algorithm

Results:

Data after using a minimum spanning tree through 39 iterations of FDGA. Clusters for all pictures are using cosine distance and have 100 clusters.

Data using 100 nearest neighbors through 10 iterations of FDGA.

Data using 10 iterations of FDGA with 100 nearest neighbors, then 1 iteration using minimum spanning tree.