Asymptotic information loss from grouping Created 2026-09-24 Updated 2026-11-05
Suppose are independent and identically distributed with an exponential distribution of rate , but only the bin indices are observed, for known bin width . These are geometric random variables on the nonnegative integers, withWhen , maximizing the grouped likelihood gives the rate maximum-likelihood estimatorIf all bin indices are zero, the likelihood increases toward its supremum as , so there is no finite rate maximizer.
Using , the central limit theorem and delta method giveFor the ungrouped observations, has limiting variance on this scale. The grouped limiting variance equals , which exceeds for positive . Thus grouping reduces asymptotic relative efficiency by the ratio . This ratio tends to one as : sufficiently fine bins lose little information, whereas a fixed bin width causes a genuine persistent precision loss. Setting recovers .
ATPase dwell-time inference from concentration dependence Created 2026-10-24 Updated 2026-11-05
A rotary ATP synthase subunit can exhibit an ATP-dependent exponential distribution of binding waits with rate and an ATP-independent sum of two first-order waits. Convolution gives the latter density . The equal-rate limit is the Erlang distribution . These fits support one concentration-dependent binding event and two kinetically resolved transitions, but do not uniquely assign the latter to hydrolysis, product release or conformational change. Their rates cannot be ordered from this symmetric density alone.
Bayes estimator under weighted absolute loss Created 2026-10-14 Updated 2026-11-05
For loss , assume , and . The posterior expected loss is a positive constant times absolute-error loss under the posterior distribution reweighted by . Thus its minimizers are the medians of that weighted posterior distribution. More precisely, the weighted cumulative distribution function satisfies ; the one-sided slopes of the convex loss give this condition.
For relative absolute loss on positive parameters, . With a gamma distribution posterior of shape two and rate , reweighting produces an exponential distribution of rate , so the estimate is . The weighted first-moment condition ensures finite loss; without it every finite action can have infinite loss, making a median characterization uninformative.
Beta-prime distribution Created 2026-10-14 Updated 2026-11-05
For , this probability density function is obtained from a beta distribution variable by . Indeed and , and the change of variables formula gives the displayed density. In particular, the ratio of two independent equal-rate exponential distributions has , density and cumulative distribution function .
Bootstrap failure for a uniform endpoint Created 2026-10-14 Updated 2026-10-18
For a continuous uniform distribution on , the scaled maximum gap tends to a unit-rate exponential distribution. An ordinary bootstrap sample includes the unique observed maximum with probability . The bootstrap maximum gap therefore retains an atom at zero, whereas the target limit is continuous. This proves failure of the ordinary empirical bootstrap for this endpoint functional; consistency of the endpoint estimate alone does not justify resampling its extreme-value limit.
Boundary bias of a kernel density estimator 2026-10-14
If the unknown probability distribution is supported on , an ordinary Gaussian density kernel still puts some fitted probability below zero. For positive observations and the standard normal distribution cumulative distribution function , the displayed formula gives the total leakage. For the statistical sample and smoothing bandwidth , it isThus about of the fitted probability falls outside the known support of a probability distribution, even though the full integral remains one. The leakage panel in smoothing bandwidth shades this area. Simply deleting the negative part leaves area ; renormalizing it makes another probability density function but does not by itself establish that the bias of an estimator has been corrected.
The issue also appears in the expectation of the kernel density estimator. For a nonnegative symmetric kernel for density estimation and a probability density function that is zero for and has a finite positive right-hand limit , assume is bounded. Then, for fixed as ,Half the averaging window lies on the impossible side of the boundary. For example, the rate-one exponential distribution has , so the limiting expected boundary height is , rather than the right-hand height one. This refers to estimating the right-hand limit; changing a probability density function at a single point does not change its probability distribution. Decreasing the smoothing bandwidth narrows the affected region but does not remove this boundary-height error. This is why the interior second-order pointwise bias bound for kernel density estimation cannot be applied across a jump at the boundary.
Change-of-variables formula for a probability density Created 2026-09-28 Updated 2026-11-05
Let a random vector have probability density function on an open set . If is a continuously differentiable diffeomorphism onto an open set , with nonsingular Jacobian matrix, then has probability density functionand zero off . A diffeomorphism includes differentiability of the inverse; a merely differentiable bijection is not the full hypothesis needed for this density formula. The determinant corrects local volume, and its absolute value accounts for an orientation reversal.
Indeed, for a Borel set , apply the change of variables formula to . The resulting integral over identifies the probability density function.
For example, if has an exponential distribution of rate and , use the diffeomorphism . It gives for . Its integrable singularity at zero does not create a point mass. A map with several inverse branches requires summing their density contributions rather than choosing a single inverse.
Compensator of a counting process Created 2026-10-05 Updated 2026-11-05
A Compensator of a counting process is a predictable increasing process , normalized by , such that is a local martingale. Under the standard local-integrability and usual-filtration conditions, the predictable compensator is unique up to indistinguishability. It describes accumulated conditional event rate, not an event probability.
For a rate- Poisson process, and is a martingale. If is exponential of rate and counts a single event, its predictable intensity is andThe rate shuts off after the event. This pathwise accumulated intensity can exceed one even before an event occurs; it differs from the deterministic probability .
A compensator need not be absolutely continuous in time. The deterministic count has compensator , because its jump time is known in advance and . No ordinary Rate function integrated against produces that step. When , the predictable derivative is the counting-process intensity. For an observed survival event under independent censoring it takes the familiar form .
Competing end-of-life criteria 2026-11-05
Competing end-of-life criteria are two or more conditions, any one of which can terminate useful operation under the stated specification. They define a remaining useful life by the earliest relevant crossing. If the prospective crossing times, measured from the present, are , thenThis is the event structure studied by competing risks. A condition ending one operating cycle must not inadvertently be treated as the component's aging endpoint: NASA's battery experiment separately defines a terminal-potential discharge cutoff and a30% rated-capacity-loss aging criterion.[ref]
Li-ion Battery Aging Datasets by NASA. Dataset description, discharge termination versus aging end-of-life threshold
For an original mechanical example, suppose a component can reach either an unacceptable vibration level in12h or an unacceptable temperature in20h under a specified operating scenario. If either condition stops operation, its predicted useful life is12h. Requiring both conditions would instead define another stopping rule. An observed early crossing also does not establish when the other criterion would have been reached if the test had continued.
For uncertain crossing times, the survival function of useful life isOnly under independence does this equal the product of the individual survival functions. For an explicitly hypothetical independent pair of exponential distributions with rates and , the product is when is measured in hours. The expected first crossing is , smaller than either individual mean,100h or50h.
The independence assumption is substantive. Shared temperature, load or damage can correlate crossing times, so multiplying marginal survival functions can give the wrong answer. Moreover, an intervention triggered by the first crossing can change the subsequent trajectory. The physical end-of-life criterion, dependence model and assumed future operation therefore belong with any numerical lifetime prediction.
The elapsed-time transition distribution of a driftless square-root diffusion starting at is the compound Poisson distributionwhere all variables on the right are independent and the empty sum is zero. Here the exponential distribution parameter is its rate. The Laplace transform of this sum iswhich is the Laplace transform of the driftless square-root diffusion; uniqueness of the Laplace transform proves the distributional identity. In particular, for , the exponential moment isFor this exponential moment is infinite when , since the event has positive probability and one exponential jump already has an infinite moment there. At the process remains zero and every such moment is one.
Conditional split of two exponential times Created 2026-10-14 Updated 2026-11-05
For independent exponential distributions with rates , let . Dividing the joint density of by the hypoexponential distribution density of gives a truncated exponential distribution on with rate . Its displayed conditional expectation supplies an E-step when only pair totals are observed. At equal rates the conditional split is uniform, with mean .
// Target: probability-and-statistics.bigb
Confidence interval for an inverse rate Created 2026-10-07 Updated 2026-11-05
For an exponential distribution with positive rate , the mean is . A confidence interval with for the rate transforms to one for the mean by reversing and inverting its endpoints. The monotonic transformation preserves the coverage event; it does not give a prediction interval for one realized holding time. A zero lower rate endpoint corresponds to an infinite upper mean endpoint.
Continuous-time Markov chain Created 2026-09-24 Updated 2026-11-05
A Continuous-time Markov chain on a finite or countable state space is specified by transition rates for , with finite total exit rate . At a state with , it waits an exponential time of rate and then moves to with probability . A state with is absorbing: it has infinite holding time rather than a zero-rate exponential waiting law.
Its transition-rate matrix is , with off-diagonal entries and diagonal entries , so each row sums to zero. The rates describe first-order transition probabilities: for . On a finite state space, the transition semigroup is and explosion is impossible. On a countably infinite state space, finite individual exit rates do not alone exclude infinitely many jumps in finite time; nonexplosion must be checked separately.
For two states with rates and , the mean holding times are and . Detailed balance gives Stationary distribution , andThis starts at one and relaxes to the stationary mass of state zero. Changing holding rates can preserve an embedded jump chain while changing the proportions of elapsed time spent in its states.
Cox–Snell residual Created 2026-10-06 Updated 2026-11-05
A Cox–Snell residual is the fitted cumulative hazard function evaluated at an observed event or censoring time: . Its reference probability distribution comes from the probability integral transform. If the true event time has a proper continuous survival function , then satisfiesFor example, an exponential distribution with rate has , so rescaling its event time by produces a unit-rate exponential distribution.
With right censoring, and the transformed observation is together with its original event indicator . Under independent censoring, a Kaplan–Meier estimator applied to these transformed observations should approximately follow ; its logarithm should approximately follow a straight line of slope minus one. Censored residuals must remain censored in this comparison.
The exact exponential identity uses the true model, whereas is estimated from data: fitted residuals are approximate diagnostics and need not be independent exact exponentials. Atoms in a discrete event-time probability distribution, or a cure fraction with positive probability of no finite event, invalidate the unmodified proper-continuous reference law. Such models require diagnostics respecting their jumps or surviving mass.
Cumulative hazard function Created 2026-09-24 Updated 2026-10-14
Suppose has an absolutely continuous probability distribution, with survival function . On an interval where , its cumulative hazard isHere is the hazard function; its integral is dimensionless because the units of rate and time cancel. The identity follows by integrating the derivative of , so mere continuity of the survival function is not enough: the absolute continuity assumption supplies a probability density function. If reaches zero, interpret as . The normalization also matters; an initial probability mass at zero would require an initial term .
Accumulating hazard functions adds on the scale and multiplies survival factors on the scale. For and ,For instance, a cumulative hazard increment of gives event probability , rather than a probability of . With per year, and . With the increasing hazard function per year when is expressed in years, , so gives . Its coefficient has units of inverse years squared.
The constant hazard function produces an exponential distribution; the increasing example is a Weibull distribution. Their survival functions both decrease, although one hazard function stays constant and the other increases. At year ten both have and despite different earlier histories. The accumulation identity is described in Lu Tian's Stanford notes, Section 1.1.
For a discrete hazard at each possible event time , the exact counterpart is . Taking its negative logarithm sums , rather than . Therefore exponentiating a Nelson–Aalen estimator does not give exactly the Kaplan–Meier estimator; small-jump comparison of cumulative hazard estimators quantifies the difference.
Delayed neutron-capture coincidence window 2026-10-18
Why not accept every delayed pulse? A longer window captures more genuine neutron capture signals, but also admits more accidental coincidences. Consider an original model with a capture waiting time following an exponential distribution of mean . Assume detection after a prompt signal accepts , with perfect time measurement and no lower cut. Integrating the normalized density givesThis assumes that moderation, competing capture channels, boundaries and electronic recovery can be represented or neglected in a single effective waiting-time model. It is not a fitted historical capture law.
Let independent prompt-like and delayed-like background streams have Poisson process rates and . The accidental coincidence pair rate in a directed window of full width is . Electronics producing at most one accepted delayed match per prompt instead give . These definitions coincide to first order when . Correlated backgrounds, such as one external neutron producing both a prompt recoil and a later capture, require separate controls; an independent-stream estimate cannot remove them.
For chosen and , the accepted fraction is . With chosen and , the pair rate is , while the at-most-one rate is approximately . These are teaching parameters, not reported Savannah River rates. A measured uncorrelated background can be tested using displaced time windows; a reactor-off measurement addresses a different question and cannot by itself characterize every reactor-correlated background.
Dependence in Poisson thinning 2026-10-14
Independent marking is the hypothesis that guarantees the Poisson thinning theorem. Equal marginal retention probabilities do not guarantee it. Consider one Bernoulli random variable with success probability , independent of a Poisson process, and retain every arrival when and none when . Each arrival is retained with probability , but the marks share the same coin. In an interval whose original count has Poisson distribution of mean , the retained count is and the rejected count is , soThe extra variance shows that is not Poisson distributed; the negative covariance proves dependence. Its no-arrival probability is , rather than . For the toy values , both rules have retained mean , but the common-coin rule has variance and no-arrival probability , compared with variance and no-arrival probability for independent marking. These formulas follow by conditioning on and using the law of total variance.
An arrival-dependent or adversarial retention rule can change the law just as strongly. Keeping every second arrival gives . Its successive retained gaps are sums of two independent exponential random variables, hence have the Erlang distribution of shape two and rate , not an exponential distribution. The resulting renewal process and its count variance are analyzed in Every second Poisson arrival has underdispersed counts, and the two dependent streams in alternating split of a Poisson process. As an even more restrictive rule, keeping only the first arrival ever gives , a Bernoulli distribution at each , so it cannot have a positive-mean Poisson distribution. A thinning rule must specify how decisions depend on arrivals and on one another; the independent-marking conclusion cannot be applied automatically to arbitrary deletion.
Equal expected recoveries from excess of loss and stop loss Created 2026-10-14 Updated 2026-11-05
For independent exponential claims of mean and a zero-based geometric count of parameter , excess of loss reinsurance at per-claim reinsurance retention has expected recovery . The zero-based geometric sum of exponential variables has tail , so aggregate stop loss reinsurance at reinsurance retention has expected recovery . Equating them gives the displayed reinsurance retention relation. This comparison concerns claim payouts, before any reinsurance premium is charged.
Equilibrium residual-life distribution Created 2026-09-24 Updated 2026-11-05
The equilibrium residual-life distribution describes the remaining wait to a renewal when observing a stationary renewal process at an ordinary time. For positive interarrivals with finite mean , it has survival function and densityThe density integrates to one by the Tail-sum formula for expectation. It exists even when the original lifetime law has atoms. A deterministic lifetime gives a uniform residual on ; an exponential lifetime gives the same exponential distribution as the original lifetime.
For the discrete chain convention in residual lifetime Markov chain, the invariant probabilities are instead for . A deterministic integer lifetime gives a uniform law on . These formulas describe observation at equilibrium; they do not make the first lifetime of an ordinary renewal process already have that residual law.
Estimation of a beta shape parameter 2026-10-25
For an independent statistical sample from the beta distribution with parameters , , the probability density function is on . Its raw moments are for , by direct integration. Consequently the method of moments gives , whereas the maximum-likelihood estimator is .
The second formula follows by setting the score function to zero; its derivative is negative. Also has an exponential distribution of rate , since . The central limit theorem and delta method giveThus the asymptotic relative efficiency of relative to is . The transformed sum has a gamma distribution of shape and rate , which also gives for and for .

