Exercise 1: Vector or Matrix? — Possible Solution ==================================================================== (A) AN RGB COLOUR — VECTOR ------------------------------ An RGB colour is a single ordered list of three numbers (red, green, blue), with no separate rows-and-columns structure — that's exactly what a vector is. It's a 3-vector in "colour space," the same way a 3D position is a 3-vector in physical space. (B) A ROTATION APPLIED TO ANY POINT IN A 2D SCENE — MATRIX ------------------------------ This isn't a single piece of data to store — it's a *rule* that takes any input vector and produces a rotated output vector. Something that transforms vectors, rather than being a vector itself, is a matrix. The same 2x2 rotation matrix can be applied to every point in the scene. (C) A SINGLE DATA POINT WITH FIVE NUMERIC FEATURES — VECTOR ------------------------------ Five numbers describing one thing (one data point) is an ordered list, not a transformation rule — a 5-dimensional vector, often literally called a "feature vector" in machine learning. (D) A FULL BLACK-AND-WHITE IMAGE OF PIXEL BRIGHTNESS VALUES — MATRIX ------------------------------ An image has a genuine two-dimensional row/column structure — a brightness value at row i, column j. That row-and-column arrangement of numbers is precisely a matrix, not a flat list. WHY THIS WORKS AS AN ANSWER ------------------------------ Each classification comes from asking the same question this chapter's own comparison table asks: is this a single ordered list of numbers (a vector), or is it either a genuinely two-dimensional grid of numbers or a rule for transforming other vectors (a matrix) — rather than guessing from the name alone.