---
title: "Greek Alphabet Reference"
subtitle: "Karena Selalu Lupa Apa Itu ψ"
---
## Complete Greek Alphabet
All 24 Greek letters, both lowercase and uppercase (where uppercase has a distinct typographical form from the Latin alphabet), with pronunciation and standard usage in statistics, econometrics, and machine learning.
| Letter | Name | Pronunciation | Common Usage in Stats / Econometrics / ML |
|--------|------|---------------|------------------------------------------|
| $\alpha$ / $A$ | Alpha | AL-fuh | Significance level ($\alpha = 0.05$); regularization strength (Ridge/LASSO); learning rate (some notations); shape parameter (Gamma, Beta distributions); Type I error rate |
| $\beta$ / $B$ | Beta | BAY-tuh | Regression coefficients ($\beta_1, \ldots, \beta_k$); beta in CAPM ($\beta = \text{Cov}(R_i, R_m)/\text{Var}(R_m)$); shape parameter (Beta distribution); Type II error probability |
| $\gamma$ / $\Gamma$ | Gamma | GAM-uh | Discount factor in economics/RL ($\gamma \in (0,1)$); Euler-Mascheroni constant ($\gamma \approx 0.5772$); Gamma function $\Gamma(n) = (n-1)!$; VAR coefficient matrices |
| $\delta$ / $\Delta$ | Delta | DEL-tuh | Change or difference ($\Delta y = y_t - y_{t-1}$); Dirac delta function $\delta(x)$; small perturbation; first difference operator in time series; treatment effect |
| $\epsilon$ / $\varepsilon$ / $E$ | Epsilon | EP-sil-on | Error term in regression ($\varepsilon_i = y_i - x_i^T\beta$); arbitrarily small positive number; $\varepsilon$-insensitive loss (SVR); machine epsilon |
| $\zeta$ / $Z$ | Zeta | ZAY-tuh | Riemann zeta function $\zeta(s) = \sum_{n=1}^\infty n^{-s}$; damping ratio; rarely used as a free parameter |
| $\eta$ / $H$ | Eta | AY-tuh | Learning rate in gradient descent ($\eta$ or $\eta_t$); canonical parameter in GLMs; efficiency ratio; $\eta^2$ (eta-squared, effect size in ANOVA) |
| $\theta$ / $\Theta$ | Theta | THAY-tuh | Generic parameter vector ($\theta \in \Theta$); angle; parameter in maximum likelihood $\hat{\theta}_{MLE}$; threshold; temperature in softmax |
| $\iota$ / $I$ | Iota | eye-OH-tuh | Vector of ones $\iota_n = (1, 1, \ldots, 1)^T$; rarely used otherwise in modern notation |
| $\kappa$ / $K$ | Kappa | KAP-uh | Condition number of a matrix ($\kappa(A) = \sigma_{max}/\sigma_{min}$); Cohen's kappa (inter-rater agreement); curvature |
| $\lambda$ / $\Lambda$ | Lambda | LAM-duh | Eigenvalue ($Av = \lambda v$); Lagrange multiplier; regularization parameter (Ridge: $\lambda\|\beta\|^2$, LASSO: $\lambda\|\beta\|_1$); hazard rate in survival analysis; Poisson rate; diagonal matrix of eigenvalues $\Lambda$ |
| $\mu$ / $M$ | Mu | MYOO | Population mean $\mu = E[X]$; mean vector $\boldsymbol\mu$ (multivariate normal); drift term; location parameter |
| $\nu$ / $N$ | Nu | NYOO | Degrees of freedom ($t_\nu$, $\chi^2_\nu$); innovation/shock in VAR models; frequency; kinematic viscosity |
| $\xi$ / $\Xi$ | Xi | ZAI or KSEE | Random variable (general); slack variable in SVM ($\xi_i \geq 0$); basis functions; auxiliary variable |
| $\pi$ / $\Pi$ | Pi | PIE | Probability (especially for Bernoulli: $\pi = P(Y=1)$); policy in reinforcement learning $\pi(a\|s)$; product operator $\Pi_{i=1}^n$; mathematical constant $\pi \approx 3.14159$ |
| $\rho$ / $P$ | Rho | ROH | Correlation coefficient $\rho = \text{Cor}(X,Y)$; spatial autocorrelation parameter; AR(1) coefficient; discount rate; spectral radius |
| $\sigma$ / $\Sigma$ | Sigma | SIG-muh | Standard deviation $\sigma = \sqrt{\text{Var}(X)}$; sigmoid function $\sigma(x) = 1/(1+e^{-x})$; singular value; $\Sigma$: covariance matrix; $\Sigma$: summation operator $\sum_{i=1}^n$ |
| $\tau$ / $T$ | Tau | TAW | Quantile in quantile regression ($\tau \in (0,1)$); time index; precision ($\tau = 1/\sigma^2$ in Bayesian); time constant; Kendall's $\tau$ |
| $\upsilon$ / $\Upsilon$ | Upsilon | UP-sil-on | Rarely used in statistics; sometimes used as an auxiliary variable |
| $\phi$ / $\Phi$ | Phi | FEE or FIE | AR coefficient in ARMA/ARIMA models; standard normal PDF $\phi(x) = \frac{1}{\sqrt{2\pi}}e^{-x^2/2}$; $\Phi(x)$: standard normal CDF; basis function; feature map in kernel methods; autoregressive polynomial |
| $\chi$ / $X$ | Chi | KIE | Chi-squared distribution $\chi^2(k)$; $\chi^2$ test statistic; characteristic function notation (sometimes) |
| $\psi$ / $\Psi$ | Psi | PSEE or SIE | Wavelet functions (mother wavelet $\psi$); digamma function $\psi(x) = \frac{d}{dx}\ln\Gamma(x)$; potential function; state in quantum mechanics analogy |
| $\omega$ / $\Omega$ | Omega | oh-MAY-guh | Frequency (angular frequency in time series); weight vector in neural networks $\omega$; $\Omega$: covariance matrix of error terms in GLS ($\varepsilon \sim N(0, \sigma^2\Omega)$); sample space in probability |
::: {.callout-note}
**Uppercase forms:** Many uppercase Greek letters are visually identical to Latin capitals (e.g., $A$, $B$, $E$, $H$, $I$, $K$, $M$, $N$, $O$, $P$, $T$, $X$, $Z$) and are therefore rarely used as separate mathematical symbols. The uppercase forms that appear in math are primarily: $\Gamma$, $\Delta$, $\Theta$, $\Lambda$, $\Xi$, $\Pi$, $\Sigma$, $\Phi$, $\Psi$, $\Omega$.
:::
---
## Variant Forms
Some Greek letters have common variant forms used in different contexts:
| Standard | Variant | Notes |
|----------|---------|-------|
| $\epsilon$ | $\varepsilon$ | $\varepsilon$ is more common in econometrics for the error term |
| $\phi$ | $\varphi$ | $\varphi$ is sometimes used for basis functions or alternative to $\phi$ |
| $\theta$ | $\vartheta$ | Rarely used; $\theta$ is standard |
| $\pi$ | $\varpi$ | Extremely rare in statistics |
| $\rho$ | $\varrho$ | Extremely rare |
| $\sigma$ | $\varsigma$ | Not used in statistics |
| $\kappa$ | $\varkappa$ | Rarely used |
---
## Common Mathematical Operators and Symbols
### Set and Logic Operators
| Symbol | Name | Meaning / Usage |
|--------|------|-----------------|
| $\in$ | element of | $x \in \mathcal{X}$: $x$ is an element of set $\mathcal{X}$ |
| $\notin$ | not element of | $x \notin \mathcal{X}$: $x$ is not in $\mathcal{X}$ |
| $\subset$ | proper subset | $A \subset B$: every element of $A$ is in $B$, and $A \neq B$ |
| $\subseteq$ | subset or equal | $A \subseteq B$: every element of $A$ is in $B$ (allows equality) |
| $\cup$ | union | $A \cup B$: elements in $A$ or $B$ or both |
| $\cap$ | intersection | $A \cap B$: elements in both $A$ and $B$ |
| $\setminus$ | set difference | $A \setminus B$: elements in $A$ but not in $B$ |
| $\emptyset$ | empty set | The set with no elements |
| $\forall$ | for all | $\forall x \in \mathcal{X}$: for every $x$ in $\mathcal{X}$ |
| $\exists$ | there exists | $\exists x$: there exists an $x$ |
| $\nexists$ | does not exist | $\nexists x$: no such $x$ exists |
| $\Rightarrow$ | implies | $A \Rightarrow B$: if $A$ then $B$ |
| $\Leftarrow$ | implied by | $A \Leftarrow B$: $A$ if $B$ |
| $\iff$ | if and only if | $A \iff B$: $A$ implies $B$ and $B$ implies $A$ |
| $\neg$ | negation | $\neg A$: not $A$ |
### Relations
| Symbol | Name | Meaning / Usage |
|--------|------|-----------------|
| $\sim$ | distributed as | $X \sim N(\mu, \sigma^2)$: $X$ follows the normal distribution |
| $\approx$ | approximately equal | $e \approx 2.718$ |
| $\propto$ | proportional to | $p(\theta \mid x) \propto p(x \mid \theta) p(\theta)$ |
| $\equiv$ | identically equal | True for all values; $\sin^2\theta + \cos^2\theta \equiv 1$ |
| $\triangleq$ or $:=$ | defined as | $\hat{\beta} \triangleq (X^TX)^{-1}X^Ty$ |
| $\perp$ | perpendicular / independent | $X \perp Y$: $X$ and $Y$ are independent; $u \perp v$: vectors are orthogonal |
| $\perp\!\!\!\perp$ | statistically independent | $X \perp\!\!\!\perp Y$: $X$ and $Y$ are independent (stronger notation) |
| $\ll$ | much less than | $\lambda \ll 1$: $\lambda$ is negligibly small |
| $\gg$ | much greater than | $n \gg k$: many more observations than parameters |
| $\leq$, $\geq$ | less/greater than or equal | Standard ordering |
| $\preceq$, $\succeq$ | matrix ordering | $A \preceq B$: $B - A$ is positive semidefinite |
### Calculus and Analysis
| Symbol | Name | Meaning / Usage |
|--------|------|-----------------|
| $\nabla$ | nabla / del | Gradient operator: $\nabla f = \left(\frac{\partial f}{\partial x_1}, \ldots, \frac{\partial f}{\partial x_n}\right)^T$ |
| $\nabla^2$ | Laplacian / Hessian | $\nabla^2 f$: Hessian matrix of second partial derivatives |
| $\partial$ | partial derivative | $\frac{\partial f}{\partial x_i}$: partial derivative of $f$ w.r.t. $x_i$ |
| $\int$ | integral | $\int_a^b f(x)\,dx$: definite integral |
| $\iint$ | double integral | $\iint_{\mathcal{D}} f(x,y)\,dx\,dy$ |
| $\sum$ | summation | $\sum_{i=1}^n x_i = x_1 + x_2 + \cdots + x_n$ |
| $\prod$ | product | $\prod_{i=1}^n x_i = x_1 \cdot x_2 \cdots x_n$ |
| $\lim$ | limit | $\lim_{n \to \infty} \bar{X}_n = \mu$ |
| $\sup$ | supremum | Least upper bound; used in analysis |
| $\inf$ | infimum | Greatest lower bound |
| $\text{argmin}$ | argument of minimum | $\hat{\theta} = \text{argmin}_\theta \mathcal{L}(\theta)$ |
| $\text{argmax}$ | argument of maximum | $\hat{\theta}_{MLE} = \text{argmax}_\theta \ell(\theta)$ |
### Norms and Inner Products
| Symbol | Name | Meaning / Usage |
|--------|------|-----------------|
| $\|\cdot\|$ or $\|\cdot\|_2$ | Euclidean norm | $\|\mathbf{x}\| = \sqrt{\sum_i x_i^2}$ |
| $\|\cdot\|_1$ | $L^1$ norm | $\|\mathbf{x}\|_1 = \sum_i |x_i|$; LASSO penalty |
| $\|\cdot\|_p$ | $L^p$ norm | $\|\mathbf{x}\|_p = \left(\sum_i |x_i|^p\right)^{1/p}$ |
| $\|\cdot\|_\infty$ | infinity norm | $\|\mathbf{x}\|_\infty = \max_i |x_i|$ |
| $\|\cdot\|_F$ | Frobenius norm | $\|A\|_F = \sqrt{\text{tr}(A^TA)}$; for matrices |
| $\langle \cdot, \cdot \rangle$ | inner product | $\langle \mathbf{x}, \mathbf{y} \rangle = \mathbf{x}^T\mathbf{y} = \sum_i x_i y_i$ |
| $|\cdot|$ | absolute value / determinant | $|x|$: abs value of scalar; $|A|$ or $\det(A)$: determinant |
### Number Sets
| Symbol | Name | Contents |
|--------|------|----------|
| $\mathbb{N}$ | Natural numbers | $\{0, 1, 2, 3, \ldots\}$ or $\{1, 2, 3, \ldots\}$ (convention varies) |
| $\mathbb{Z}$ | Integers | $\{\ldots, -2, -1, 0, 1, 2, \ldots\}$ |
| $\mathbb{Q}$ | Rationals | Fractions $p/q$ with $p, q \in \mathbb{Z}$, $q \neq 0$ |
| $\mathbb{R}$ | Real numbers | All real numbers; $\mathbb{R}^n$: $n$-dimensional real vector space |
| $\mathbb{R}_+$ | Positive reals | $\{x \in \mathbb{R} : x > 0\}$ |
| $\mathbb{R}_{++}$ | Strictly positive reals | Same as $\mathbb{R}_+$ in some notations |
| $\mathbb{C}$ | Complex numbers | $\{a + bi : a, b \in \mathbb{R}\}$ |
| $\mathbb{R}^{m \times n}$ | Real matrices | Space of $m \times n$ real matrices |
| $\mathbb{S}^n$ | Symmetric matrices | $\{A \in \mathbb{R}^{n\times n} : A = A^T\}$ |
| $\mathbb{S}^n_+$ | PSD matrices | $\{A \in \mathbb{S}^n : A \succeq 0\}$ |
### Typographic Conventions
| Convention | Meaning | Examples |
|------------|---------|---------|
| Lowercase italic | Scalar | $x$, $\alpha$, $n$, $\lambda$ |
| Lowercase bold | Column vector | $\mathbf{x}$, $\boldsymbol{\beta}$, $\boldsymbol{\mu}$ |
| Uppercase (or uppercase bold) | Matrix | $X$, $A$, $\Sigma$, $\mathbf{W}$ |
| Uppercase italic | Random variable | $X$, $Y$, $Z$ |
| Lowercase italic | Realization of r.v. | $x$, $y$, $z$ |
| Calligraphic | Set or space | $\mathcal{X}$, $\mathcal{H}$, $\mathcal{L}$, $\mathcal{F}$ |
| Blackboard bold | Number set or expectation | $\mathbb{R}$, $\mathbb{E}[X]$, $\mathbb{P}(A)$ |
| Hat ($\hat{\phantom{x}}$) | Estimator or fitted value | $\hat{\beta}$, $\hat{y}$, $\hat{\sigma}^2$ |
| Tilde ($\tilde{\phantom{x}}$) | Alternative estimator | $\tilde{\beta}$ (e.g., GLS vs OLS) |
| Bar ($\bar{\phantom{x}}$) | Sample mean / average | $\bar{x} = n^{-1}\sum x_i$ |
| Star ($^*$) | Optimal or true value | $\theta^*$, $x^*$ |
---
*Back to [Appendix Overview](index.qmd) | Next: [Common Proofs](common-proofs.qmd)*