Chap 3 - Common Univariate Random Variables

Uniform, Bernoulli, Binomial and Poisson Distributions

The Uniform Distribution

The Bernoulli Distribution

The continuous uniform Distribution

The Binominal Distribution

The Poisson Distribution

Normal and Lognormal Distributions

The Normal Distribution

The Lognormal Distribution

The normal Distribution

The Standard Normal Distribution (Z-Distribution)

Calculating probabilities using z-values

Students's T, Chi-Squared, F-Disttributions

Student's t-Distribution

The Chi-Squared Distribution

The F-Distribution

defined over a range (a ,b)

has 2 possible outcomes (failure success)

used for assessing the probability of binary outcomes

n independent Bernoulli trials

defines the probability of x success in n trials

The confidence interval (khoảng tin cậy)

Range of values around E[x]

chính xác trong 1 khoảng thời gian

mean: 0; standard deviation: 1

z-value: độ lệch chuẩn giữa 1 giá trị quan sát (observation) và population mean

F(Z)

skewed to the right

Definition

Small samples (n<30); unknown variance

2 parameters: Mean, variance (standard deviation)

Key properties

X: the number of successes per unit

λ : the average number of successes per unit

used for hypothesis tests concerning the variance of a normal distributed population

used for hypothesis tests concerning the quality of the variances of 2 populations

X ~ N(μ, σ2)

Skewness = 0

Kurtosis = 3

Similar to Normal Distribution; fatter tails

Properties

symmetrical

defined by df; df = number of sample -1 = (n - 1)

fatter tails than normal distribution (greater probaility in the tails)

df increases => t-distribution looks more and more like to standard normal distribution

defined by 1 parameter

asymetrical

bounded below by 0

Others

Mixture Distributions

right-skewed

being truncated at zero on the left-hand side

defined by 2 separate df

Key Properties

bounded by 0 (always >=0)