Continuous Distributions Updated (1)
Continuous Random Variables
Continuous random variables differ from discrete variables in that they can take any value over an interval, such as (a,b), (-∞, 0), (0,∞), and (-∞,∞).
The probability distribution associated with a continuous random variable X is known as the probability density function (pdf), denoted by f(x).
Events are defined as the area under the density curve for specific ranges of X, making cumulative distribution functions (cdf) meaningful: P(X ≤ x).
Key Concepts
Probability Density Function (PDF) and Cumulative Density Function (CDF)
Applications of Continuous Random Variables
Commonly modeled variables in business include:
Weights of products (cans, packages).
Delivery and completion times (projects, packages).
Financial metrics (returns, costs).
Rates (hotel occupancy, churn, etc.).
Continuous Probability Distributions
Characteristics of Density Curves
Each density curve indicates the likelihood of all possible outcomes, where the total area under the curve equals 1.
For any single value of X, the probability is zero: P(X = c) = 0.
Probabilities are discussed for intervals (regions).
Probability Density Function Requirements
f(x) ≥ 0 for every x in [a,b].
The total area under the curve from a to b equals 1.
To find the probability of a range (c ≤ X ≤ d), compute the area under the graph of f(x) from c to d.
Uniform Distribution
A continuous random variable X follows a uniform distribution, denoted X ~ U(a,b), if it has a constant pdf over the interval.
Formula:
Example - Uniform Distribution
For X ~ Uniform(2, 6), pdf is:
f(x) = 0.25 for 2 ≤ x ≤ 6.
Probability that the bus is late between 3 and 5 minutes:
P(3 ≤ X ≤ 5) = (5 - 3) / (6 - 2) = 0.5.
Normal Distribution
Overview
The normal distribution is a key continuous distribution characterized by its bell shape.
Represents many real-world phenomena.
Defined by mean (μ) and variance (σ²).
Notable properties:
Symmetric, with mean = median = mode.
Range: (-∞, +∞).
Approximately 68% of data falls within ±1σ, 95% within ±2σ, and 99.7% within ±3σ.
Transformation to Standard Normal Distribution (Z)
Any normally distributed variable can be standardized to Z:
Z = (X - μ) / σ.
The standard normal distribution has a mean of 0 and standard deviation of 1.
Finding Normal Probabilities
To calculate the probability P(a < X < b), Standardize X to Z-value and use the Z-table to find probabilities:
Example:
For a normally distributed X with μ = 18 and σ = 5, to find P(X < 18.6):
Z = (18.6 - 18) / 5 = 0.12.
Use cumulative distribution tables or statistical software for calculations.
Empirical Rule
The Empirical Rule approximates the dispersion of data in a normal distribution:
About 68% data falls in the range μ ± 1σ.
About 95% data falls in the range μ ± 2σ.
About 99.7% data falls in the range μ ± 3σ.
Exponential Distribution
Definition
A right-skewed distribution used to model time between events (e.g., waiting times).
Key parameters include λ (lambda), which represents the rate of occurrence.
Key Properties
The pdf for an exponential variable:
f(x) = λe^{-λx}, for x ≥ 0.
Mean and standard deviation are both equal to 1/λ.
Example
If customers arrive at a rate of 15 per hour, the average inter-arrival time is:
P(T < 3 minutes) = 1 - e^{-λx} for λ = 15/60 and x = 3/60.
Binomial to Normal Approximation
As the number of trials n increases, Binomial distributions can be approximated as Normal if np ≥ 10 and n(1 - p) ≥ 10.
Example Calculations:
For n=100 and p=0.5, μ = np = 50, σ = sqrt(np(1-p)) = 5.
These notes provide a comprehensive overview of continuous random variables, including key distributions like uniform, normal, and exponential, their properties, and applications in statistical modeling.