Maths for ML / Notation

\mathbb{R}^{n}, \mathbb{R}^{m\times n}

real vector / matrix space

Number systemsBeginner

Say it “R n / R m by n”

ℝⁿ is the set of length-n real vectors; ℝ^{m×n} is the set of m-by-n real matrices. It tells you shape.

Example

W \in \mathbb{R}^{768\times 3072}

W is a 768 × 3072 matrix.

In code

W.shape == (768, 3072)

W.shape == torch.Size([768, 3072])

Where you’ll see it in ML

The fastest way to learn a tensor's shape from a paper.

Don’t confuse it with

$\mathbb{R}$ real numbers All numbers on the number line, including decimals and irrationals.
$\in$ element of The thing on the left is a member of the set on the right.

Lessons that use it

← All 94 symbols