Traditional Graph Analysis Methods

Data Science, AI and ML
#1

Traditional methods are mostly algorithm-based, such as :

  1. searching algorithms, e.g. BFS, DFS
  2. shortest path algorithms, e.g. Dijkstra’s algorithm, Nearest Neighbour
  3. spanning-tree algorithms, e.g. Prim’s algorithm
  4. clustering methods, e.g. Highly Connected Components, k-mean

The limitation of such algorithms is that we need to gain prior knowledge of the graph at certain confidence before we can apply the algorithm. In other words, it provides no mean for us to study the graph itself. And most importantly, there is no way to perform graph level classification.