Problem of old evidence 2026-10-24
Already accepted evidence can support a newly considered theory even though conditioning again on a fact of probability one produces no increase in its posterior probability. Distinguish new learning, retrospective support and discovering a previously unrecognized theory–evidence connection.[ref]
Product measure Created 2026-09-24 Updated 2026-11-05
For sigma-finite measure spaces and , the product measure is the unique measure on the product sigma-algebra satisfying
Here is interpreted as zero. Sigma-finiteness supplies the usual existence and uniqueness theorem. For Lebesgue measures on two real lines, the product is planar Lebesgue measure and the formula assigns a rectangle the product of its side lengths. Tonelli's theorem identifies integration of nonnegative measurable functions with iterated integration; Fubini's theorem gives the integrable signed version.
For a countable family of probability spaces , the product probability measure is characterized by finite-coordinate cylinder sets:
This constructs independent coordinate random variables with the specified marginal laws. The infinite-coordinate statement uses probability measures; it is not an unrestricted existence claim for arbitrary infinite products of measures of infinite mass.
Product probability space 2026-10-14
For finitely many probability spaces , their product probability space has underlying set , the product sigma-algebra and product measure . The coordinate projections are jointly independent: for measurable , the probability of all coordinate conditions is . Conversely a family of independent coordinates has this product joint distribution. Product structure permits coordinate exposure and independent resampling in concentration arguments.
Product stationarity under single-coordinate Markov updates Created 2026-10-14 Updated 2026-11-05
Let be a Markov kernel with Stationary distribution on one coordinate. On coordinates, update coordinate with kernel while leaving the others unchanged. The product distribution is invariant: its updated coordinate still has law , and independent transition randomness leaves it independent of the untouched coordinates. Choosing the updated coordinate independently of the current configuration, with any fixed selection probabilities, forms a mixture of these invariant kernels and preserves the same product law.
For a two-state kernel with and , the stationary Probability of state one solves . If , then
Under the product stationary distribution, the occupied-coordinate count has binomial distribution with parameters , mean and variance . In the exceptional case , the kernel is the identity and every is stationary; the displayed fraction is undefined and must not be used.
State-dependent coordinate selection can invalidate the conclusion. On two bits, let each single-coordinate kernel refresh to a fair bit, but select coordinate one whenever its current value is zero and coordinate two otherwise. Starting with independent fair bits, the Probability that the first bit equals one after a step is , so the uniform product law is not invariant. Invariance of each coordinate-update kernel alone is insufficient when the mixture weights depend on the state.
Projective measurement Created 2026-09-28 Updated 2026-10-14
A projective measurement has self-adjoint orthogonal projections satisfying and . Measuring an observable with discrete spectral decomposition reports eigenvalue with Born rule probability . A projector can have rank greater than one: the outcome then identifies an eigenspace, rather than a particular eigenvector inside it. Projectors specify the outcome statistics; the usual ideal conditional update is the additional Lüders rule.
In the qubit example from measurement in quantum mechanics, . At , the zero and one probabilities are and , regardless of the relative quantum phase . Under the Lüders rule, obtaining zero prepares ; obtaining one prepares , up to global phase. Repeating the same measurement returns the same label with certainty, provided no intervening evolution or operation changes the measured sector. Repeatability concerns the conditional output, and does not mean the first outcome was predetermined.
Change to and its Pauli X eigenstates . Adding the two probability amplitudes before squaring gives
At and , and . At , both are ; at , the first pair is exchanged. The statistics remain throughout. This is the same addition-and-cancellation principle as wave interference: relative quantum phase changes how amplitudes reinforce or cancel in another basis.
The beauty of this basis dependence is that information invisible to one measurement becomes accessible to another. A measurement of an ensemble estimates its quantum state populations, while probes one component of its quantum coherence. These are experiments on separate identically prepared copies. Sequentially measuring and then on one copy instead probes the disturbed post-measurement state. Quantum state tomography combines suitable ensemble statistics to reconstruct the original state.
For labelled pure states in dimension , an orthonormal basis supplies error-free projective state exclusion when for every . The Born rule gives zero probability to reporting on input ; orthonormal-basis completeness ensures that an outcome is always obtained. Conversely, zero error for this labelled basic measurement forces precisely those zero overlaps. The matrix of transition probabilities is therefore a matrix of column probability distributions with zero diagonal.
Propensity score Created 2026-09-28 Updated 2026-11-05
The propensity score for binary treatment and pretreatment covariates is . It is a balancing score:
This distributional fact does not require conditional exchangeability of the potential outcomes. It follows from the definition of the true assignment probability.[ref]
For a bounded function , condition successively on and on :
Also , so division by gives the same conditional covariate distribution among treated units as in the full score stratum; the analogous argument uses for controls. Comparison of both arms requires .
Identifying a causal effect is an additional step. If , then the balancing property lets the scalar propensity score replace the full measured in that exchangeability statement. Without this causal assumption, balancing measured covariates leaves possible unmeasured confounding. An estimated score also need not deliver exact finite-sample covariate balance; its estimation and the subsequent matching in causal inference or weighting procedure must be checked.
Proper mean lifetime 2026-10-25
The mean lifetime of an unstable particle measured in its rest frame, using elapsed proper time. In the exponential decay model, survival to proper time has probability . For constant Lorentz factor , time dilation gives laboratory mean lifetime . A mean lifetime is a statistical mean, rather than a fixed decay time.
Proper scoring rule Created 2026-10-07 Updated 2026-10-14
A proper scoring rule makes reporting the true probability distribution optimal in expected value, for every true probability distribution in the allowed class of forecasts. In the reward convention, the inequality in the title holds for every report and true law . The expected value is always taken under , because the report does not change how the outcome is generated. In the loss convention the same definition reads
This requirement makes a proper scoring rule useful for eliciting probabilities and comparing predictive distributions: deliberately distorting a known forecast cannot improve its expected reward or expected loss. The reward convention and its strict version are set out in Strictly Proper Scoring Rules, Prediction, and Estimation by Tilmann Gneiting and others (Section 1).
For the Bernoulli distribution with , the Brier score has expected loss , and the negative logarithmic scoring rule has expected loss . Both are minimized at the honest report . The linear probability score, by contrast, maximizes its expected reward at . These are statements about averaging over both possible outcomes with their actual probabilities.
The left panel displays each excess expected loss divided by that rule's expected loss at : for the Brier score and approximately for the logarithmic scoring rule. This normalization permits comparison of curve shapes and preserves the minimizing report, since it subtracts a constant and divides by a positive constant independent of . Zero means no excess expected loss. The right panel shows the unmodified linear probability score reward, with its maximizing direction labeled explicitly.
Property B 2026-10-14
A hypergraph has Property B if its vertices of a hypergraph can be coloured with two colours so that every hypergraph edge meets both colour classes. An -uniform hypergraph with fewer than edges has Property B, by a fair independent colouring and the union bound. Conversely, for , take vertices and independently sample uniformly random -subsets. In any fixed colouring an edge is monochromatic with probability at least : convexity of binomial coefficients reduces to the balanced colouring, and . The union bound over colourings gives a probability at most of any proper colouring. Removing repeated edges leaves a hypergraph without Property B and with edges.
For positive total effort, a proportional allocation contest awards player the prize with probability , or gives it that proportion of a divisible prize. A convention is needed at zero total effort. This allocation is different from the discontinuous highest-effort rule of an all-pay auction.
Protein sequence motif 2026-11-05
A protein sequence motif is a characteristic pattern of amino acid residues in a protein sequence. A pattern can specify fixed residues, alternatives and variable-length or wildcard positions. It is a statement about residue order; it need not specify a complete protein structural motif or independently establish biochemical function. PROSITE documents such signatures together with their interpretation and matching syntax.[ref]
For an original example, use the invented four-residue pattern D-x(2)-[ST]: an initial D, any two residues, then S or T. With 20 possible standard amino acids, it admits
strings out of possible four-residue strings. Under an explicitly stipulated uniform independent-residue model, a given window therefore matches with probability . This is not an empirical background model for real proteins, and the pattern is not asserted to be a real biological signature.
A sequence of length 103 has 100 length-four windows. Under that model the expected value of the number of matches is . Overlapping windows need not be independent for this expectation: summing their indicator expectations is sufficient. Thus finding a short pattern can occur without a demonstrated functional relationship.
A useful annotation combines the pattern with the rest of the sequence, structural context and experimental evidence. In a calcium-binding family, for example, an EF hand depends on the spatial arrangement and chemistry of coordinating groups, not merely on a count of pattern matches.[ref]
Pseudorandom function family Created 2026-09-28 Updated 2026-11-05
A pseudorandom function family is a collection indexed by a key , with an efficient evaluation algorithm. For a random key, an adversary receiving answers to chosen queries should be unable to distinguish from a uniformly random function on the same finite domain and range, within a declared resource bound.
Both experiments answer repeated queries consistently. Fresh queries to the random function receive independent uniform values; queries to use one fixed key throughout. The adversary may choose its next query using previous answers.
For a one-bit output, an all-zero family is easily distinguished: two different queries return with probability one there and probability for a random function. State key length, query bound, running time and distinguishing advantage when using the definition quantitatively.[ref]
Pseudorandom number Created 2026-10-14 Updated 2026-10-20
A pseudorandom number is an output of a deterministic algorithm chosen to imitate samples from a specified probability law. For example, a linear congruential generator returns fractions on a finite grid in . Their intended uniform distribution is an approximation, and their serial dependence must be assessed separately from the generator's period.
Pure-investment arbitrage 2026-10-05
A pure-investment arbitrage has zero initial cost, no intervening consumption, and a nonnegative terminal payoff that is strictly positive with positive probability.
Pure-state gentle measurement bound Created 2026-10-05 Updated 2026-11-05
If is a positive contraction and , the normalized successful post-measurement state is . Since ,
A likely outcome therefore preserves a pure input well in quantum fidelity when its measurement operator is positive. Positivity excludes a hidden unitary rotation that could otherwise change the state even for an outcome of probability one.
For an orthonormal basis and nonnegative numbers summing to one, the displayed vector is a unit vector and gives measurement probabilities by the Born rule. Conversely, every pure state with those probabilities has coefficients of these magnitudes, with arbitrary phases on the nonzero coefficients. A common change of all phases is a global phase; differences between them are relative phases that can affect measurements in another basis. Zero-probability coefficients vanish, so their phases have no effect.
For two normalized pure states, the difference of their projectors has two eigenvalues , giving that trace distance. Write a mixed density operator as . Convexity of the trace norm and weighted Cauchy-Schwarz inequality give . Thus a small probability of failing the pure-state projector test forces trace-distance proximity, including for correlated multipartite outputs.
Purity of a density operator Created 2026-09-24 Updated 2026-10-18
The purity of a density operator is . It equals one exactly for a pure state; in finite dimension it lies between and . The minimum occurs exactly at the maximally mixed state .
A worked qubit family interpolates between the mixed state and the pure state , where :
The eigenvectors are , with eigenvalues . Squaring and summing these eigenvalues gives
Here is the Born rule probability of the positive outcome in an projective measurement. The Bloch vector is , since with one of the Pauli matrices. At the purity is ; at it is , with ; at it is one. Within this family, is a pure state and every is a mixed state. Extending the formula to produces a negative eigenvalue , so it ceases to be a density operator despite remaining Hermitian with trace one.
Figure 1.
Purity and X-measurement probability as a qubit moves from the Bloch-ball centre to a pure state
. Details.
The Bloch vector length measures polarization: for any qubit density operator, . This connects the interior of the Bloch ball with mixed states and its boundary, the Bloch sphere, with pure states. [ref]
For a short proof of the general bound, use the spectral theorem: the nonnegative eigenvalues sum to one, so . Then
The lower bound is attained only when all eigenvalues equal ; the upper bound is attained only when exactly one eigenvalue is nonzero, and hence equals one. Purity is therefore a simple quantitative test of mixedness, invariant under a unitary operator. It does not specify the entire density operator: even distinct pure states all have purity one. [ref]