A larger baryon-to-photon ratio favours bound nuclei at a fixed temperature through the equilibrium factor . In particular it helps deuterium survive photodissociation sooner: there are fewer energetic photons per baryon, and nuclear reaction rates per baryon also rise. A schematic deuterium bottleneck condition is , suppressing slowly varying powers of temperature. It gives
Thus higher permits burning at a higher temperature. During radiation domination that means earlier cosmic time, , leaving less time for free-neutron decay after freeze-out. More neutrons enter helium, so the final helium-4 mass fraction increases, ordinarily rather weakly. This is the baryon-density effect on primordial helium.
The equilibrium helium factor indicates the direction of the change but is not a prediction that the final helium abundance grows without bound as . Once almost all surviving neutrons are captured, the final primordial helium mass fraction is limited by their supply, as the shared neutron budget shows.
A longer neutron half-life means a larger mean lifetime . Holding the initial neutron-to-proton ratio and the nuclear-burning delay fixed, the derivative of the neutron survival factor is positive:
Thus fewer neutrons disappear during the deuterium bottleneck, and the final helium-4 mass fraction increases.
There can be a second effect if the increased lifetime represents a uniform reduction of the relevant weak interaction strength. With fixed masses and other couplings, neutron decay and the conversion reactions both scale as . Thus , whereas . Weaker conversion then freezes out earlier and increases , reinforcing the survival effect. The survival argument alone already establishes the requested neutron half-life effect on primordial helium; linking lifetime to all conversion rates additionally requires this microscopic assumption.
Extra relativistic neutrino species increase the effective number of relativistic energy degrees of freedom. At fixed temperature the Hubble parameter rises as , while the standard conversion rate is unchanged. Consequently cosmological weak freeze-out moves to higher temperature, , and the neutron-to-proton ratio increases. The initial neutron fraction therefore increases as well.
The quicker expansion also reduces the cosmic time required to cool to the deuterium bottleneck temperature: during radiation domination, while the thermal count is roughly constant. Fewer neutrons decay before burning. Both factors in the neutron budget increase, giving a larger final helium-4 mass fraction. This is the primordial helium response to extra relativistic species, under the assumption that the additional species contribute mainly to expansion rather than independently changing the conversion rates.
For actual final density operator and target , minimize
This Hilbert-Schmidt distance is nonnegative and is zero exactly at the target, for pure or mixed states. The subscript denotes evolution under the chosen control; is fixed. A simple overlap need not have its maximum at a general mixed target, whereas this distance has the required equality condition. For purely Quantum Hamiltonian dynamics, the target must lie on the initial state's unitary orbit of a density operator to attain zero error; otherwise the constrained optimization finds the closest accessible state instead.
For target unitary operator and actual propagator on an -dimensional space, minimize the phase-insensitive unitary gate error
The Cauchy-Schwarz inequality for the Hilbert-Schmidt inner product gives a value in , because both unitary matrices have squared norm . The error is zero exactly when the matrices are proportional, namely . It therefore tests the full process, not just its action on one state, and ignores a physically irrelevant common phase. If the gate phase itself must be fixed, instead minimize , which vanishes only at exact equality.
For a normalized state driven by the admissible control , maximize
Here is the specified Hermitian operator, the final time, and the dependence on is through the Schrödinger equation. The objective is real and is exactly the desired expectation value, so increasing it increases the requested observable. A control-resource penalty may be included when the admissible set does not already constrain field strength or fluence.
The adiabatic dark state of a driven Hamiltonian has exactly zero component on the lossy intermediate level throughout the ideal passage. It therefore suppresses spontaneous emission, whereas the two-pulse route explicitly transfers the population into before moving it to . For an intermediate-level decay rate , the leading loss probability is controlled by . In stimulated Raman adiabatic passage, nonadiabatic leakage makes of order , rather than the order-one intermediate population of the sequential swaps. Transfer is also less sensitive to precise pulse areas, provided the ordering, gap and slowness requirements hold. Making the process slower indefinitely is not always beneficial if the outer states also decohere; their coherence must survive the total passage duration.
For real nonnegative envelopes, put and choose a continuous . The state is dark because its two amplitudes feeding cancel: . The orthogonal bright combination couples to with strength . Thus the other eigenstates are with energies .
Start with and , so and . Slowly increase to infinity while maintaining a nonzero gap; end with and , so . This target-to-intermediate pulse first is the counterintuitive ordering of stimulated Raman adiabatic passage. Its quantitative slowness condition is
It follows from and the gap . Avoid switching both envelopes off while is changing; switch off together only after the transfer angle has become stationary. The dark eigenvalue is zero, so there is no dynamical phase; with this real eigenvector as well. The ideal adiabatic quantum control therefore transfers to , physically the same target ray.
Using the EM for merged multinomial cells E-step,
Applying the M-step maximum-likelihood estimator to the filled counts gives
Initialize . At each iteration first compute by the conditional expectation above. The M-step applies the complete-data maximum-likelihood estimator to the filled counts, giving
Repeat until convergence. If all counts are zero, the likelihood function is constant and any parameter can be retained. For a nonempty sample an interior start makes the displayed denominator positive.
Here the three observed cell probabilities are , and , so the observed log-likelihood is
For a nonempty sample its second derivative on is
Thus the maximum-likelihood estimator is unique, allowing a boundary maximum. To verify that the iteration actually finds it, let and . Then , and
If the maximum is interior, this identity moves each iterate towards it, and monotonicity of prevents crossing it. The bounded monotone iterates converge; continuity makes their limit a fixed point, and strict concavity makes that fixed point the maximum. If the maximum is at zero or one, the sign of gives the same monotone convergence towards the appropriate endpoint. In the special case , is identically one whenever its denominator is positive, so the maximum is reached in one step. An interior initialization avoids the possible spurious zero fixed point when ; starting at zero can otherwise trap the iteration even when the true maximum is positive.
Within the merged third observed cell, the probability of belonging to the original fourth cell is
The allocation therefore has the conditional binomial distribution
Its conditional expectation and the filled-in missing counts are
with and .
Taking the conditional expectation of the complete log-likelihood gives the EM for merged multinomial cells E-step:
This is the parameter-dependent part of . More precisely, , where the constant is independent of the candidate , interpreting factorials of noninteger filled counts through the gamma function. The expected log-factorials are not generally the log-factorials of expected counts, so literal equality including the normalization would be false. Equality up to this constant is exactly what the M-step needs. Filling in conditional means works here because the candidate-parameter terms are linear in the missing counts; it is not a universal rule for the expectation-maximization algorithm.
The parameter-dependent part of the complete multinomial distribution log-likelihood is
The remaining probability factors and multinomial coefficient are constant in . If , differentiating gives , whose zero is
The log-likelihood is concave. If , its maximum is at zero; if , its maximum is at one. These boundary cases are included by the same formula. If , the data lie entirely in the parameter-independent fourth cell and every is a maximum-likelihood estimator.
Reversible-jump Markov chain Monte Carlo requires the reciprocal density/proposal ratio for the matched reverse transformation, so the death acceptance probability is
Expressed directly in the current residual and the recovered lower-order parameters, this is
Indeed makes this exponent the negative of the birth exponent, and all prefactors invert. Thus for matched states. This reciprocity, including the normalized prior distributions and auxiliary proposal density, gives the required detailed balance between models.
For the reverse centered polynomial birth move in reversible-jump sampling, recover from the current coefficient vector. Remove that highest-order coefficient and undo the intercept shift:
No auxiliary random number is drawn for this death move: the removed coefficient supplies the auxiliary value needed to reconstruct the birth move. The reverse Jacobian determinant is also one. The inverse prediction change is , hence the residual in the lower-order model is .
A reversible-jump Markov chain Monte Carlo calculation compares normalized model-specific densities. The fixed-order posterior written only up to proportionality in part (a) omits normalizing factors that depend on and therefore cannot simply be canceled between dimensions. Moreover, the printed question does not specify the model-order prior distribution or the probabilities of selecting birth and death moves. Denote those by , and , respectively, retaining them in the answer.
Put , , , and . The centered polynomial birth move in reversible-jump sampling changes the fitted vector by , so its new residual is . Therefore
The map from to has ones down the diagonal and a single additional entry in the intercept row. Its Jacobian determinant is one.
Write and . The normalized multivariate normal distribution prior ratio is
Divide also by the forward proposal density . The common inverse-gamma distribution prior for and the common normal-likelihood normalization cancel, since stays fixed. The forward acceptance ratio is consequently
Here is the proposal standard deviation. If equal model priors and symmetric birth/death selection are explicitly adopted, their ratios become one; they are not determined by the PDF. The expression assumes the same hyperparameters for the variance prior in both models, as in part (a). A dimension-dependent variance prior would contribute its own ratio.
Choose an initial . At each iteration, evaluate and using and draw from the conditional multivariate normal distribution just derived. Next compute and draw
For example, generate and set . A draw of the conditional multivariate normal distribution can be formed as with a Cholesky decomposition and a vector of independent standard normal distributions.
This two-block Gibbs sampler alternates the full conditional distributions of the independent normal and inverse-gamma regression priors posterior. The conditional-integration argument proves that its invariant distribution is the joint posterior distribution. Under the usual convergence conditions, retain the pairs after an initial transient as the requested dependent posterior sample. The new variance update uses the newly drawn coefficients, rather than the previous coefficients.
Use the shape-scale convention for the inverse-gamma distribution: has density proportional to for . Write , , , and . The independent normal and inverse-gamma regression priors make the conditional log-likelihood plus log prior in equal, up to a constant, to
Define and . Completing the square turns this expression into plus a constant. Thus the conditional posterior distribution is the multivariate normal distribution
Positive definiteness of the prior covariance matrix makes this well defined even if the design matrix is rank deficient.
Holding fixed, the factors depending on are
Consequently
There is no additional in the shape: the normal distribution prior for is independent of . A prior scaled by would be a different model and would change that calculation.
A systematic Gibbs sampler updates the coordinates of successively from their full conditional distributions. At iteration , draw
For a single coordinate, if the current state has distribution , its untouched coordinates have the correct marginal law and the resampled coordinate has precisely the conditional law under . Integrating marginal times conditional therefore recovers . Each coordinate kernel preserves this invariant distribution, so their composition also preserves it. This proves the target of the Gibbs sampler; the whole systematic sweep need not be a reversible kernel.
Starting from a suitable state and assuming the required irreducibility and convergence conditions, discard an initial transient and retain subsequent states as a dependent sample from, or approximately from, . In finite-state problems irreducibility and aperiodicity suffice; in continuous-state problems the corresponding positive recurrence and ergodicity conditions must hold. Successive draws are generally dependent, so Monte Carlo uncertainty must account for that dependence.
For Beta sampling by gamma ratios, use either preceding chi-squared distribution algorithm to generate independent and , and return
Write , so and . The Jacobian determinant is . The joint probability density function after this change of variables factors as
Integrating out proves the Beta distribution for .
A new method is Beta sampling by uniform order statistics. Generate independent uniforms, sort them, and return the th order statistic. There are uniforms below the selected value and above it. Selecting their labels and placing the selected observation in gives density
which is the Beta distribution density. The integrality of the shape parameters is what makes this order-statistic construction possible.
The first algorithm uses independent normal distributions: generate with law and return
For example, the Box-Muller transform generates two such normal variates from independent uniforms:
To verify this, has density and is an independent uniform angle. Converting the resulting polar density to Cartesian coordinates gives , which factors into two independent normal distributions. Independent pairs therefore give the required sum, discarding a spare coordinate when is odd. Its moment-generating function is , that of the chi-squared distribution.
A genuinely different algorithm uses exponential-envelope rejection sampling for a chi-squared variable. Put . The target and proposal probability density functions are
For , differentiating gives , so the maximum is at . The envelope constant and normalized density ratio are
They remain valid at , where and . Repeatedly draw fresh independent , put , and accept when
The right side is nonpositive by , so this is a legitimate acceptance probability. The joint probability of proposing and accepting in is ; conditional on acceptance the density is exactly . This proves the rejection sampling algorithm, rather than only identifying the target law.
For I prefer the second method: it reduces to a single exponential draw with no rejection. The first method is simple and attractive when normal draws are already available or is small. For larger , the second method avoids generating normals: its acceptance probability is , asymptotic to by Stirling's formula. Its expected number of proposals grows like , rather than the linear number of normal draws in the first method. Both are exact ideal-uniform algorithms; actual runtime also depends on the implementation.

Pinned article: Introduction to the OurBigBook Project

Welcome to the OurBigBook Project! Our goal is to create the perfect publishing platform for STEM subjects, and get university-level students to write the best free STEM tutorials ever.
Everyone is welcome to create an account and play with the site: ourbigbook.com/go/register. We belive that students themselves can write amazing tutorials, but teachers are welcome too. You can write about anything you want, it doesn't have to be STEM or even educational. Silly test content is very welcome and you won't be penalized in any way. Just keep it legal!
We have two killer features:
  1. topics: topics group articles by different users with the same title, e.g. here is the topic for the "Fundamental Theorem of Calculus" ourbigbook.com/go/topic/fundamental-theorem-of-calculus
    Articles of different users are sorted by upvote within each article page. This feature is a bit like:
    • a Wikipedia where each user can have their own version of each article
    • a Q&A website like Stack Overflow, where multiple people can give their views on a given topic, and the best ones are sorted by upvote. Except you don't need to wait for someone to ask first, and any topic goes, no matter how narrow or broad
    This feature makes it possible for readers to find better explanations of any topic created by other writers. And it allows writers to create an explanation in a place that readers might actually find it.
    Figure 1.
    Screenshot of the "Derivative" topic page
    . View it live at: ourbigbook.com/go/topic/derivative
  2. local editing: you can store all your personal knowledge base content locally in a plaintext markup format that can be edited locally and published either:
    This way you can be sure that even if OurBigBook.com were to go down one day (which we have no plans to do as it is quite cheap to host!), your content will still be perfectly readable as a static site.
    Figure 5. . You can also edit articles on the Web editor without installing anything locally.
    Video 3.
    Edit locally and publish demo
    . Source. This shows editing OurBigBook Markup and publishing it using the Visual Studio Code extension.
  3. https://raw.githubusercontent.com/ourbigbook/ourbigbook-media/master/feature/x/hilbert-space-arrow.png
  4. Infinitely deep tables of contents:
    Figure 6.
    Dynamic article tree with infinitely deep table of contents
    .
    Descendant pages can also show up as toplevel e.g.: ourbigbook.com/cirosantilli/chordate-subclade
All our software is open source and hosted at: github.com/ourbigbook/ourbigbook
Further documentation can be found at: docs.ourbigbook.com
Feel free to reach our to us for any help or suggestions: docs.ourbigbook.com/#contact