Graph

What is a graph

What is graph-based learning

An Affinity Matrix, also called a Similarity Matrix, is an essential statistical technique used to organize the mutual similarities between a set of data points. … These similarity measures can be interpreted as the probability that that two points are related.Representing GraphsThe following sections detail the matrices used to represent the graphs processed in this experiment. The term-document data is represented as a graph G = (V,E), where V = v1, v2, … , vn is the vertex set. Each vertex vi represents a document in the dataset. The edge set E indicates a connection between vertices. In this case, an edge would represent a shared term.

Why represent data as a graph

https://youtu.be/RS5xoGE2yZ4?t=3474 we’ll go over in one of our lectures another reason so you can cluster data more easily from a graph representation for sure another reason is you can understand paths of progression trajectories or differentiation more easily basically the idea is you take a walk on a graph so if you start walking on this graph you’ll get a path through the data in tomorrow’s lecture we’ll talk about diffusion which is a probabilistic way of taking a walk through the graph and you’re going to compute probabilities of getting from one point to another that will emphasize very major pathways in the data so that’s another reason why you want to represent it as this kind of graph because every time you’re at a point you can tell what other points the edges the point is connected to by an edge and you can choose to go to the strongest edge for example with higher probability and that’ll take you on a walk through the main path in the data and that way you can understand trajectories in the data okay