確率分布

In any probability distribution, each individual probability must fall between zero and one, and all probabilities must sum to exactly one. These constraints ensure the distribution is mathematically valid and accounts for all possible outcomes of a random variable. This foundation applies across all distribution types.

A binomial probability distribution applies to scenarios with a fixed number of trials where each trial has exactly two outcomes, like coin tosses. A Poisson distribution models independent events occurring within specific intervals, such as website visitors per hour. Binomial requires predetermined trial counts; Poisson focuses on event frequency over time or space.

A uniform distribution represents probabilities evenly spread across all possible values within a range, creating a rectangular shape. Examples include voltage from an electric company or equal likelihood of drawing any suit from a card deck. This contrasts with distributions where some outcomes are more probable than others.

The normal distribution forms a symmetric bell-shaped curve and describes many real-world phenomena like birth weights and IQ scores. Its predictable shape allows statisticians to estimate probabilities and make inferences about populations. However, it applies only to certain types of data and cannot be universally applied.

A Poisson distribution approximates a binomial distribution when the probability of success is small (≤0.05) and the number of trials is large (≥20). This approximation simplifies calculations while maintaining reasonable accuracy. The conditions ensure the binomial distribution's shape resembles the Poisson model closely enough for practical use.

Probability distributions divide into discrete and continuous types. Discrete distributions apply to countable outcomes like coin flips or website visitors, subdividing into binomial and Poisson types. Continuous distributions apply to measurable values across ranges, subdividing into uniform and normal types, each serving different analytical purposes.

Probability distributions are commonly expressed as formulas, graphs, or tables, each providing different insights into the data. Formulas enable precise calculations, graphs visualize probability patterns, and tables organize discrete probabilities systematically. The choice of representation depends on the distribution type and analytical context.