Positive association of two events Created 2026-10-06 Updated 2026-11-05
For nontrivial events, strict attraction under conditional probability is equivalent to a positive covariance of their indicator variables. This condition is symmetric in the two events. Complementing exactly one of the events reverses the strict inequality. This two-event condition should be distinguished from association of an entire family of random variables.
Positive difference of independent exponential waiting times Created 2026-10-18 Updated 2026-11-05
Let be independent random variables with exponential distributions of respective positive rates . Condition on and use the survival function of :At this is . Dividing by that probability proves that the positive difference conditional on again has an exponential distribution with rate . The memorylessness of the exponential distribution persists after an independent competing event has already occurred. Replacing the difference by a negative sentinel when adds an atom of a measure of mass at that sentinel without changing the positive tail.
Power of a two-sample rare-event comparison Created 2026-10-14 Updated 2026-11-05
For two independent samples of size with event probabilities , put , , and . The normal approximation to a two-sided level- test has powerFor rare events, independent Poisson distributions approximate the counts. Conditional on their total, the null count allocation is binomial distribution with probability one half; under the alternative the allocation probability is . Averaging its exact rejection probability over the alternative total supplies a discrete power calculation.
// Target: probability-theory.bigb
Predictable compensator of a jump measure Created 2026-10-14 Updated 2026-11-05
The predictable random measure matching the expected integrals of the jump measure against every nonnegative predictable integrand. For displacement marks of a Markov jump process, . Multiplying the integrand by an event known before an interval proves that the compensated integral has conditionally centered increments; expected absolute integrability makes it a true martingale. With the compensator absolutely continuous in time, squared-integrand integrability gives the square-integrable compensation isometry.
Probability Created 2026-10-03 Updated 2026-10-14
The probability of an event is its value under a probability measure, a real number between zero and one. It compares the mass assigned to that event with the unit mass of the whole sample space. For a finite sample space with equally likely outcomes, . For unequal outcome masses, add their individual probabilities instead.
For example, selecting a coin with head probability with chance , or a coin with head probability with chance , gives head probability . This is the law of total probability: averaging over the hidden selection is part of calculating the outcome's probability.
Probability inequality Created 2026-09-24 Updated 2026-10-18
Probability inequalities bound event probabilities from limited information, such as moments, without requiring a complete probability distribution. Markov's inequality uses a nonnegative random variable's expected value; Chebyshev's inequality uses a variance to bound deviations from a mean. The attraction is a guarantee that applies to many possible probability distributions. Their bounds can be conservative for a particular model, and distributions attaining equality show when the available information really cannot give more.
Probability mass function Created 2026-09-24 Updated 2026-11-05
For a discrete probability distribution, the probability mass function of its random variable isThe sum ranges over its finite or countable set of possible values. An event's probability is the sum of the masses of the included values. For the example under probability distribution, . The cumulative distribution function sums masses below a threshold; its jump at a value equals that value's probability mass. This differs from a probability density function, whose height is not a point probability.
Probability measure Created 2026-09-24 Updated 2026-10-14
A probability measure is a nonnegative, countably additive measure with total mass one. Thus , and for pairwise disjoint events ,The complement of an event has probability . The probability of a union of overlapping events needs a correction: . Counting the overlap once is the simplest instance of the inclusion-exclusion principle.
A zero-probability event need not be impossible as a member of the sample space. Under the continuous uniform distribution on , every individual point has probability zero, while the whole interval has probability one. Countable additivity applies to countable unions, not arbitrary uncountable ones; there is no contradiction.
Probability model Created 2026-10-25 Updated 2026-11-05
A mathematical account of random outcomes using a probability space and specified random variables. It determines probabilities for events or observed data. A statistical model commonly supplies a family of such probability distributions with unknown parameters; choosing a model and estimating its parameters are distinct tasks.
Probability space Created 2026-09-24 Updated 2026-10-14
A probability space is a measure space whose total measure is one. The sample space contains the elementary outcomes; the sigma-algebra contains the events to which the model assigns probabilities; the probability measure supplies those values. The sigma-algebra lets us combine observable questions using complements and countable unions while remaining within the model.
For two independent uniform dice, take , , and give each ordered pair mass . A random variable might record the sum, the larger result, or whether both results are even integers. These are different functions on the same sample space, rather than different elementary experiments. A probability distribution records the mass of the values of one such random variable and can forget information retained by the full probability space.
Probability theory Created 2026-09-24 Updated 2026-10-18
Probability theory describes uncertainty through a probability space: possible outcomes form a sample space, observable questions are events, and a probability measure assigns them numbers between zero and one. The Kolmogorov axioms make those assignments consistent. They let us combine information about several events, update it through conditional probability, and study random variables without listing every possible outcome separately.
For example, roll two dice whose results are independent random variables, each with a discrete uniform distribution on . There are 36 equally likely ordered outcomes. Only have sum seven, so that event has probability . The sum itself is not uniformly distributed: sum two has only one outcome. Choosing a probability distribution for the elementary outcomes is part of the model; equal likelihood does not follow merely from counting the named possibilities.
The subject becomes especially useful when complete outcome lists are enormous or infinite. Independence turns a joint calculation into simpler pieces, expected value summarizes a random variable, and the law of large numbers explains how averages can become predictable even when individual observations are uncertain. The central limit theorem describes the scale and shape of many remaining fluctuations. These are distinct claims: a small variance of an average does not make every observation close to its expected value.
There is a satisfying unity here: dice, hidden causes, and Brownian motion all use the same rules for events and conditional probability. A model still needs justification. Its exact deductions describe what its assumptions imply; comparison with observations belongs to statistical inference.
Queue (queueing theory) Created 2026-10-14 Updated 2026-11-05
A queue consists of customers or jobs waiting for or receiving service. An arrival process adds customers and a service rule specifies which waiting customer is served next. A stochastic model of these events is studied in queueing theory; its queue length is the number of customers waiting or in service, according to the stated convention.
Random variable Created 2026-09-24 Updated 2026-10-14
A real-valued random variable is a measurable function from a probability space to the real numbers with their Borel sigma-algebra. More general targets give random elements. For every Borel set , measurability ensures that is an event, so the question "does the observed value lie in ?" has a defined probability.
The function is fixed; its input outcome is uncertain. This separates what we measure from how the uncertainty is modeled. A probability distribution is the resulting pushforward measure, . A cumulative distribution function, probability mass function, or probability density function describes that measure in an appropriate setting. The random variable and its probability distribution are different objects: the former also retains its relationship to other measurements on the same sample space.
For two independent uniform dice, let record their results and define and . The 36 ordered pairs are equally likely, but the resulting probability mass functions differ:For the sum, count the diagonal of the outcome grid having total . For the maximum, subtract the pairs with both results below from the pairs with both results at most . These examples show how a deterministic function transforms a simple input probability distribution into a nonuniform output one.
The corresponding expected values are and . The first follows immediately from linearity of expectation; the second follows from . Neither number must be an attainable result. An expected value describes an average under the probability measure, not a prediction that the next observation will equal it.
The distinction between a random variable and its probability distribution matters for dependence. and have the same discrete uniform distribution and are independent random variables. and a second copy of also have that same marginal probability distribution, but they always agree. Their joint probability distribution is concentrated on the diagonal; the marginals alone cannot say which relationship holds. This is why a model for several measurements must specify their joint law or an appropriate independence assumption.
Indicator random variables turn events into numbers: equals one on and zero elsewhere, and . This connects event counting with expected value and variance, and lets tools for random variables answer questions about probabilities. More complicated functions of observations can describe totals, waiting times, or estimation errors using the same framework.
Rare-event collaboration size at half power Created 2026-10-14 Updated 2026-11-05
Suppose each collaborating centre contributes independently randomized participants per arm with specified event probabilities. At 50% statistical power, the usual normal-approximation alternative quantile is zero. The required size per arm is , with for a two-sided test or for a one-sided test. Divide by the per-centre arm size and round up. Endpoints must use the same time origin and follow-up window, and cluster dependence requires an additional adjustment.
Relative risk from group compositions 2026-10-14
Let be the event of membership in the first of two groups and the event of selection. Write , and , with . Bayes' theorem gives the conditional probabilities and . Their risk ratio is the displayed expression: the unknown overall selection probability cancels. Equivalently, it is the ratio of the group-membership odds after selection to those before selection.
When both selection probabilities lie strictly between zero and one, the selection odds ratio is and generally depends on . Thus compositions alone do not identify that odds ratio or provide sample sizes for a standard error. When , the first group's observed selection rate is lower; this comparison alone does not establish a causal effect of group membership.
Renewal process Created 2026-09-24 Updated 2026-11-05
A renewal process counts successive events whose interarrival times are independent and identically distributed. Allow nonnegative finite interarrival times, but require . Its renewal epochs and count areThe nondegeneracy condition makes almost surely, so the count is finite at every finite time. Zero interarrival times can give simultaneous renewals; strictly positive ones give distinct epochs. At each renewal the future interarrival sequence starts afresh, which models repeated replacement or completion events.
Exponential interarrival times give a Poisson process. Constant interarrival times give . A general renewal count is not a Markov process by itself: predicting the next arrival usually requires knowing the time since the last renewal, because a general interarrival law is not memoryless.
If , then almost surely. The strong law gives ; sandwiching between and proves the counting-rate limit. No exponential or continuous interarrival distribution is required for this long-run rate.
Residual service time 2026-11-05
A residual service time is the unserved part of a job's full-capacity service time. If the original service time is and units of service have already been delivered, its remaining amount is conditioned on the event . For ,provided the denominator is positive.
Observation during a stationary sequence of independent and identically distributed services produces a different calculation. Conditional on sampling a busy instant uniformly in service time, with , the equilibrium residual satisfiesThe second identity assumes a finite second moment. A service of deterministic duration gives a uniform distribution on and mean . Long services are encountered more often than short services, explaining why the observed residual need not resemble the original probability distribution.[ref]
Discrete Stochastic Processes (draft of 2nd edition, 31 January 2011) by Robert Gallager. Section 4.4, renewal-reward time averages and equation 4.15 for residual life
Reverse Fatou inequality for events 2026-10-25
For any sequence of events, its limsup of events satisfiesSet . These are decreasing events with intersection . Continuity of probability from above givesThis is also the bounded-indicator form of the reversed Fatou lemma. The inequality can be strict: independent fair-coin head events each have probability , while their limsup of events has probability by the Second Borel-Cantelli lemma.
Robert Kass Created 2026-10-25 Updated 2026-11-05
Robert Kass connects Bayesian statistics with statistical methods for neural data. His research includes Bayes factors, prior distributions, asymptotic inference and the analysis of spike trains. His institutional research statement describes using point-process models to analyze the timing of neuronal activity and relationships between brain regions.[ref]
https://www.stat.cmu.edu/~kass/about.html. Author Bio Sketch, Bayesian inference and neural-data analysis
A spike train is naturally represented by a point process: the data include event times rather than only a continuous response or total count. A statistical model can then express how firing intensity changes with an experimental condition or recent activity. This makes the choice of model part of the scientific question about neural communication, while uncertainty calculations distinguish a fitted pattern from a reproducible relationship.
Homepage: www.stat.cmu.edu/~kass/
Robert E. Kass
. Sample quantile Created 2026-10-14 Updated 2026-11-05
For observations and , the sample quantile in the generalized-inverse convention iswhere is the empirical distribution function and the observations are ordered with ties retained. For five observations, the sample median in this convention is the third order statistic. Interpolated sample-quantile conventions can give different finite-sample answers.
For the asymptotic statement, assume the observations are independent and identically distributed with cumulative distribution function . Let be its unique -quantile, and suppose has a density continuous and strictly positive at . ThenFor a proof, is exactly the event that at least observations are at most . At , the count is binomial with success probability . The binomial central limit theorem gives the displayed normal limit. A larger density at the target quantile makes the quantile more precisely estimated. The independence and local-density hypotheses are needed; the formula cannot be applied unchanged to arbitrary dependent observations or an atom.

