An adjacency matrix represents connections between graph nodes. Rows and columns identify nodes; each cell records whether the corresponding pair is connected. A value of 0 means no connection, while 1 means a connection exists. section 5.3.1
Use an adjacency matrix to:
- Represent an undirected graph with a symmetric matrix, because each connection works in both directions. section 5.3.1
- Prepare spectral clustering by combining the adjacency matrix with a degree matrix to calculate the Laplacian matrix. section 5.3.2
- Extend binary connections to weighted similarities, where larger edge weights indicate greater similarity between data points. section 5.3.1
Reading a connection
For nodes 1 through 5, a cell at row 1 and column 5 contains 1 when nodes 1 and 5 are connected. A cell at row 1 and column 4 contains 0 when they are not connected.
