1. What Is a Distribution?
- A probability distribution describes how the values of a random variable are spread or allocated across possible outcomes.
- It tells us:
- which values are possible,
- how likely each value is.
Mathematically:
- For a discrete random variable $X$: $P(X = x) = p(x)$
- For a continuous random variable: $P(a \leq X \leq b) = \int_a^b f(x)\,dx$ where $f(x)$ is the probability density function (PDF).
2. Key Functions of a Distribution
- PMF (Probability Mass Function) → for discrete variables (e.g., dice).
- PDF (Probability Density Function) → for continuous variables (e.g., height).
- CDF (Cumulative Distribution Function): $F(x) = P(X \leq x)$ gives probability that a variable is less than or equal to $x$.
3. Types of Distributions
Discrete Distributions
- Bernoulli: Success/failure (coin flip).
- Binomial: # of successes in $n$ trials.
- Poisson: Count of events in a fixed time/space interval.
- Geometric: # of trials until first success.
Continuous Distributions
- Uniform: All values in interval equally likely.
- Normal (Gaussian): Bell-shaped, common in nature.
- Exponential: Time until an event occurs.
- Gamma / Weibull: General lifetime distributions.
- t-distribution: Heavy-tailed, used in inference.
4. Examples
- Tossing a fair coin:
- Distribution = Bernoulli(0.5).
- Rolling a die:
- Distribution = Discrete uniform on {1,2,3,4,5,6}.
- Adult height:
- Distribution ≈ Normal($\mu=170, \sigma=10$).
- Number of daily customer arrivals:
- Distribution ≈ Poisson($\lambda=20$).
5. Why They Matter
- Distributions let us:
- Model uncertainty in data.
- Make predictions about future outcomes.
- Do statistical inference (hypothesis testing, confidence intervals).
- Perform simulations (Monte Carlo).
6. In Forecasting
- Deterministic forecast → single number.
- Probabilistic forecast → distribution of future values (e.g., “Next week’s demand follows Normal(500, σ=30)”).
- Quantiles, prediction intervals, and risk measures all come from distributions.
Summary:
A probability distribution describes how likely different values of a random variable are. Distributions can be discrete (Bernoulli, Binomial, Poisson) or continuous (Normal, Exponential, Uniform). They’re essential for quantifying uncertainty, making forecasts, and performing statistical inference.
