EE8405 Chap.2 Random Variables, Distributions and Moments
Random Variables, Distributions and Moments
This chapter turns outcomes into numerical variables and studies how probability is distributed across their values. It distinguishes mass, density, cumulative and complementary functions, then computes expectation, second moments, variance and standard deviation with their units intact.
The central queueing lesson is that two service distributions can share a mean yet create different delay because their spread and tails differ.
What this chapter covers
- 01
Probability mass
- 02
Density functions
- 03
Cumulative probability
- 04
Expectation
- 05
Variance and standard deviation
- 06
Second moments
Worked example · free
Compute a discrete service-time variance
- 1Check probability mass sums to one.
- 3Calculate first and second moments.
- 2Subtract the squared mean and retain units.
Key terms
- Second Moment
- The expectation of the squared variable, equal to variance plus squared mean.
- Density Function
- A function whose area over an interval gives probability for a continuous variable.
Random Variables, Distributions and Moments FAQ
Why is density not a point probability?
A continuous variable assigns zero probability to an exact point. Probability is area under the density over an interval, so a density value can exceed one without violating probability rules.
When do variances add?
Variances add directly for independent or zero-covariance variables. Otherwise a covariance term is required, even though expectations still add whenever the relevant means exist. State the dependence condition explicitly.
Why keep the full distribution?
Means compress location but conceal spread and tail behaviour. Queue delay, service-level breaches and capacity decisions can differ sharply between distributions with the same mean. Compare second moments as well.
How do units help audit a moment calculation?
A mean has the same unit as the variable, a second moment and variance have squared units, and standard deviation returns to the original unit after the square root. Keeping those labels beside each line exposes substitution of standard deviation for variance and prevents a second moment from being reported as spread. A final negative variance is an arithmetic warning, not a feature of the distribution.
What makes a named distribution a defensible model?
Connect its assumptions to an observable mechanism. A binomial count needs a fixed number of comparable independent trials, a Poisson count needs the process assumptions over an interval, and exponential waiting relies on memorylessness. Give one observation supporting the choice and one that would refute it. A familiar curve or convenient formula is not sufficient evidence that the distribution represents arrivals or service.
What should be checked before calculating moments?
Confirm nonnegative probabilities, unit total mass or area, and support consistent with the physical variable. These checks prevent invalid inputs from producing plausible averages. Audit them before moments and preserve the units.
Exam move
Build value-probability tables and calculate first and second moments on separate lines. Use units and a nonnegative variance as immediate error checks. Start by checking support and normalisation. A discrete mass function must be nonnegative and sum to one; a continuous density must be nonnegative and integrate to one.
The support should match the physical variable, so service time cannot be negative and a count should not take fractional values. Distinguish the mass or density from the cumulative function. At a discrete jump, greater than and greater than or equal are different events, so copy the inequality before using a complement.
For moment calculations, create columns for value, probability or density contribution, first-moment contribution and second-moment contribution. Calculate the mean and second moment independently, then obtain variance as second moment minus squared mean. Keep time-squared units through variance and return to time only after taking the square root for standard deviation.
A negative variance is an arithmetic warning, commonly caused by early rounding or subtracting the mean rather than its square. Use coefficient of variation to compare spread across scales, but translate it back through the mean before inserting variance or second moment into a queueing formula. Practise sums of random variables with and without independence.
Expectations add whenever the relevant means exist, while variances require zero covariance or an explicit covariance term. State that condition rather than silently dropping dependence. Choose named distributions from mechanisms.
Binomial needs a fixed number of common-probability independent trials, Poisson describes interval counts under process assumptions, exponential describes waiting in a Poisson process, and deterministic service has zero variance. Write one observation that supports each choice and one that would refute it. On the handwritten page, keep the variance identity, unit ladder, complement rule and covariance term together.
Finish numerical work by explaining which distribution feature drives the later queue: mean for offered load, second moment for general-service waiting, tail probability for a service breach, or memorylessness for a Markov description.
Working through Random Variables, Distributions and Moments in EE8405? Sia is AskSia’s AI Electrical Engineering tutor — ask any EE8405 Random Variables, Distributions and Moments question and get a clear, step-by-step explanation grounded in how EE8405 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
