17.3: Apéndice C- Datos sobre algunas distribuciones comunes
- Page ID
- 151097
Distribuciones discretas
Función de indicador\(X = I_E\)\(P(X = 1) = P(E) = p\)\(P(X = 0) = q = 1 - p\)
\(E[X] = p\)\(\text{Var} [X] = pq\)\(M_X (s) = q + pe^s\)\(g_X (s) = q + ps\)
Variable aleatoria simple\(X = \sum_{i = 1}^{n} t_i I_{A_i}\) (una forma primitiva)\(P(A_i) = p_i\)
\(E[X] = \sum_{i = 1}^{n} t_ip_i\)\(\text{Var} [X] = \sum_{i = 1}^{n} t_i^2 p_i q_i - 2 \sum_{i < j} t_i t_j p_i p_j\)\(M_X(s) = \sum_{i = 1}^{n} p_i e^{st_i}\)
Binomial\((n, p)\)\(X = \sum_{i = 1}^{n} I_{E_i}\) con\(\{I_{E_i} : 1 \le i \le n\}\) iid\(P(E_i) = p\)
\(P(X = k) = C(n, k) p^k q^{n - k}\)
\(E[X] = np\)\(\text{Var} [X] = npq\)\(M_X (s) = (q + pe^s)^n\)\(g_X (s) = (q + ps)^n\)
MATLAB:\(P(X = k) = \text{ibinom} (n, p, k)\)\(P(X \ge k) = \text{cbinom} (n, p, k)\)
Geométrico (\(p\))\(P(X = k) = pq^k\)\(\forall k \ge 0\)
\(E[X] = q/p\)\(\text{Var} [X] = q/p^2\)\(M_X (s) = dfrac{p}{1 - qe^s}\)\(g_X (s) = \dfrac{p}{1- qs}\)
Si\(Y - 1\) ~ geométrica\((p)\), así que\(P(Y = k) = pq^{k - 1}\)\(\forall k \ge 1\), entonces
\(E[Y] = 1/p\)\(\text{Var} [X] = q/p^2\)\(M_Y (s) = \dfrac{pe^s}{1 - qe^s}\)\(g_Y (s) = \dfrac{ps}{1 - qs}\)
Binomial negativo\((m, p)\),\(X\) es el número de fracasos antes del éxito\(m\) th.
\(P(X = k) = C(m + k - 1, m - 1) p^m q^k\)\(\forall k \ge 0\)
\(E[X] = mq/p\)\(\text{Var} [X] = mq/p^2\)\(M_X (s) = (\dfrac{p}{1 - qe^s})^m\)\(g_X (s) = (\dfrac{p}{1 - qs})^m\)
Para\(Y_m = X_m + m\), el número del juicio en el que se\(m\) produce el éxito. \(P(Y = k) = C(k - 1, m - 1) p^m q^{k - m}\)\(\forall k \ge m\).
\(E[Y] = m/p\)\(\text{Var} [Y] = mq/p^2\)\(M_Y(s) = (\dfrac{pe^s}{1 - qe^s})^m\)\(g_Y (s) = (\dfrac{ps}{1 - qs})^m\)
MATLAB:\(P(Y = k) = \text{nbinom} (m, p, k)\)
Poisson\((\mu)\). \(P(X = k) = e^{-\mu} \dfrac{\mu^k}{k!}\)\(\forall k \ge 0\)
\(E[X] = \mu\)\(\text{Var}[X] = \mu\)\(M_X (s) = e^{\mu (e^s - 1)}\)\(g_X (s) = e^{\mu (s - 1)}\)
MATLAB:\(P(X = k) = \text{ipoisson} (m, k)\)\(P(X \ge k) = \text{cpoisson} (m, k)\)
Distribuciones absolutamente continuas
Uniforme\((a, b)\)\(f_x (t) = \dfrac{1}{b - a}\)\(a < t < b\) (cero en otra parte)
\(E[X] = \dfrac{b + a}{2}\)\(\text{Var} [X] = \dfrac{(b - a)^2}{12}\)\(M_X (s) = \dfrac{e^{sb} - e^{sa}}{s(b - a)}\)
Triangular simétrico\((-a, a)\)\(f_X (t) = \begin{cases} (a + t)/a^2 & -a \le t < 0 \\ (a - t)/a^2 & 0 \le t \le a \end{cases}\)
\(E[X] = 0\)\(\text{Var} [X] = \dfrac{a^2}{6}\)\(M_X (s) = \dfrac{e^{as} + e^{-as} - 2}{a^2 s^2} = \dfrac{e^{as} - 1}{as} \cdot \dfrac{1 - e^{-as}}{as}\)
Exponencial\((\lambda)\)\(f_X(t) = \lambda e^{-\lambda t}\)\(t \ge 0\)
\(E[X] = \dfrac{1}{\lambda}\)\(\text{Var} [X] = \dfrac{1}{\lambda^2}\)\(M_X (s) = \dfrac{\lambda}{\lambda - s}\)
Gamma\((\alpha, \lambda)\)\(f_X(t) = \dfrac{\lambda^{\alpha} t^{\alpha - 1} e^{-\lambda t}}{\Gamma (\alpha)}\)\(t \ge 0\)
\(E[X] = \dfrac{\alpha}{\lambda}\)\(\text{Var} [X] = \dfrac{\alpha}{\lambda^2}\)\(M_X (s) = (\dfrac{\lambda}{\lambda - s})^{\alpha}\)
MATLAB:\(P(X \le t) = \text{gammadbn} (\alpha, \lambda, t)\)
Normal\(N(\mu, \sigma^2)f_X (t) = \dfrac{1}{\sigma \sqrt{2\pi}} \text{exp} (-\dfrac{1}{2} (\dfrac{t - \mu}{\sigma})^2)\)
\(E[X] = \mu\)\(\text{Var} [X] \sigma^2\)\(M_X (s) = \text{exp} (\dfrac{\sigma^2 s^2}{2} + \mu s)\)
MATLAB:\(P(X \le t) = \text{gaussian} (\mu, \sigma^2, t)\)
Beta\((r, s)\)
\(f_X (t) = \dfrac{\Gamma (r + s)}{\Gamma (r) \Gamma (s)} t^{r -1} (1 - t)^{s - 1}\)\(0 < t < 1\),\(r > 0\),\(s > 0\)
\(E[X] = \dfrac{r}{r + s}\)\(\text{Var} [X] = \dfrac{rs}{(r + s)^2 (r + s + 1)}\)
MATLAB:\(f_X (t) = \text{beta} (r, s, t)\)\(P(X \le t) = \text{betadbn} (r, s, t)\)
Weibull (\(\alpha, \lambda, \nu\))
\(F_X (t) = 1 - e^{-\lambda (t - \nu)^{\alpha}}\),\(\alpha > 0, \lambda >0, \nu \ge 0, t \ge \nu\)
\(E[X] = \dfrac{1}{\lambda^{1/\alpha}} \Gamma (1 + 1/\alpha) + \nu\)\(\text{Var} [X] = \dfrac{1}{\lambda^{2/\alpha}} [\Gamma (1 + 2/\lambda) - \Gamma^2 (1 + 1/\lambda)]\)
MATLAB: (\(\nu = 0\)solo)
\(f_X (t) = \text{weibull} (a, l, t)\)\(P(X \le t) = \text{weibull} (a, l, t)\)
Relación entre distribuciones gamma y Poisson
- Si\(X\) ~ gamma\((n, \lambda)\), entonces\(P(X \le t) = P(Y \ge n)\) donde\(Y\) ~ Poisson\((\lambda t)\).
- Si\(Y\) ~ Poisson\((\lambda t)\), entonces\(P(Y \ge n) = P(X \le t)\) donde\(X\) ~ gamma\((n, \lambda)\).