All MCQs

Sparse graphs contain far fewer edges than possible pairs. Adjacency lists use memory roughly proportional to vertices plus edges rather than the square of vertex count.

An adjacency matrix records pairwise connections using rows and columns for vertices. For a simple graph, entries are commonly 1 for an edge and 0 otherwise.

A weighted graph attaches a numerical strength, cost, or frequency to an edge. Message count measures the intensity of the relationship.

Following is not necessarily reciprocal, so the relationship has an origin and destination. A directed edge can represent A following B without implying B follows A.

A mutual friendship relation has no separate source and destination under the usual definition. An undirected edge represents the connection symmetrically.

Edges encode connections such as friendship, communication, following, or collaboration. They may also carry direction, weight, or other attributes.

Vertices, also called nodes, represent the objects participating in the network. Their exact meaning depends on the social system being modeled.

PCA searches only for linear directions of high variance. Weighted combinations can also obscure the direct semantic meaning of original variables.

Reversing all signs changes the orientation of the axis but not the line it spans or the variance captured. Scores reverse correspondingly, leaving reconstruction and geometry unchanged.

Retained scores are multiplied by their component directions and the feature means are restored. With fewer components, the result approximates rather than exactly reproduces the original data.

Principal components are weighted combinations and therefore transform the coordinate system. Feature selection preserves the meanings of chosen original variables.

Cumulative explained variance shows how much information is retained as components are added. A threshold offers a practical balance between compression and variance preservation.