Reading the symbols

After this lesson you can

• Name and pronounce the Greek letters ML papers use, and predict the job each one usually does
• Decode decorations ($\hat{\theta }$, $\bar{x}$, ${\theta }^{\star }$, $\tilde{B}$, $f^{\prime }$) and indices ($x_i$, ${\mathbf{x}}^{(i)}$, $W^{[l]}$, $x_t$)
• Read set, function and probability notation such as $\in$, ${\mathbb{R}}^{n\times d}$, $f:{\mathbb{R}}^n\to \mathbb{R}$, $p(y\mid \mathbf{x})$, $\mathbb{E}$ and $\sim$
• Tell $\mathrm{argmin}$ from $\min$, $\odot$ from $\circ$, and $\log$ from ${\log }_2$ — and translate each into NumPy

In this lesson

1. Why notation exists, and why it isn’t standard
2. The explorer
3. Greek letters and their usual jobs
4. Decorations: hats, bars, stars, tildes, primes
5. Indices: which one, which example, which layer, which time
6. Sets and spaces: where things live
7. Functions, definitions and near-equalities
8. Norms, inner products and the products you can confuse
9. Probability in one screen
10. Optimisation: min versus argmin, and the derivatives
11. Logs, exponentials, indicators and big-O
12. The same notation, in NumPy