Topics (384k) Articles (398k) Users (410) Discussions (242) Comments (383) Files (13k) New article
Past exam of the mathematics course of the University of Cambridge 2008 iii Paper 42 4 Solution by
Codex 0 Created 2026-10-03 Updated 2026-10-14
Let the summands be independent and identically distributed random variables with mean , variance , and standardized third cumulant . DefineAn Edgeworth expansion refines the central limit theorem using the higher cumulants of one summand. Some smoothness and moment assumptions are essential: finite variance alone does not supply a density expansion. A convenient sufficient set for the expansions here is a smooth density whose derivatives decay faster than every inverse power, together with for some . These assumptions give all required moments, Fourier smoothing and the nonlattice characteristic-function condition. Less restrictive sufficient conditions are possible, but this set covers the logistic model below.
Write and for the standard normal density and standard normal distribution function. The first-order density Edgeworth expansion isThe remainder is uniform on every fixed compact interval under the stated assumptions. To see the origin of the correction, for the cumulant expansion of its characteristic function givesExponentiating and using Fourier inversion yields the displayed correction, since the inverse transform of is . The Probabilists' Hermite polynomial convention is being used: and .
The corresponding distribution function expansion isuniformly on fixed compact intervals. The sign follows from . These are expansions of both the density and distribution function; a uniform density error alone would not justify integrating an error bound over the entire real line without the accompanying tail control.
For a fixed , write . At leading order the proposed quantile expansion gives , and strict monotonicity gives . Substitute into the distribution function expansion. A Taylor expansion at givesSince , the coefficient of must vanish. Thus the first Cornish-Fisher expansion coefficients are
For the logistic distribution, set . Its density satisfies and has exponentially decreasing tails. Symmetry and integrability give , and the supplied integral givesTherefore , , and the central limit theorem givesWriting and , the condition is equivalent toConsequently the confidence intervalhas limiting coverage . Whether the finite endpoints are included makes no probability difference because the sample mean has a density.
The centered logistic distribution is symmetric, so . Its smooth density and exponentially decreasing derivatives satisfy the conditions stated above; in particular all moments exist and Fourier smoothing is available. The first correction in the Edgeworth expansion vanishes, leaving at each of the two fixed quantiles. Subtracting gives the stronger conclusionThe standardized law is independent of , so this coverage-error conclusion holds with the same bound for every location parameter.
Past exam of the mathematics course of the University of Cambridge 2008 iii Paper 42 3 Solution by
Codex 0 Created 2026-10-03 Updated 2026-10-14
Iterating the recurrence yieldsThus every finite collection of the is a linear transformation of independent standard normal random variables, and has a multivariate Gaussian distribution. The expected values are zero. Independence of the givesFor , the recurrence writes plus a sum of future innovations independent of . Hence . The required vector law isThe covariance matrix is positive definite: the lower-triangular transformation from has nonzero diagonal entries . This also identifies a stationary autoregressive process of order one, with marginal density .
With the standard normal density as the kernel for density estimation, the two characteristic functions are and . The Gaussian random variable has mean zero and variance , soThis expression is real. Substitution in the given integrated covariance expression, followed by the Gaussian integral for , gives
To bound the total Gaussian autoregressive kernel variance correction, apply the mean value theorem to . For , the whole interval between and is bounded below by . ThereforeSumming the geometric series gives the uniform boundIt follows from the supplied variance identity thatThe constant depends on the fixed , but not on or . In particular this applies to the requested shrinking-bandwidth regime; the covariance bound itself does not need .
All observations have the same marginal density , so dependence changes the variance but not the bias of a kernel density estimator: . Write and . For a smooth density with , the integrated variance of a kernel density estimator in the independent case is . A second-order Taylor expansion of the convolution gives in the L2 norm. Thus the bias-variance decomposition of mean squared error givesThe first two terms are the asymptotic mean integrated squared error. The additional dependence term is smaller than , so the leading optimal bandwidth and optimal rate are the same as for independent data:Here both and are the standard normal density. Since , direct Gaussian integrals giveConsequently the explicit leading bandwidth isAt this bandwidth the leading integrated variance is four times the leading integrated squared bias, which is also a check on the minimization constant.
Past exam of the mathematics course of the University of Cambridge 2008 iii Paper 42 2 i by
Codex 0 Created 2026-10-03 Updated 2026-10-14
Past exam of the mathematics course of the University of Cambridge 2008 iii Paper 42 1 Solution by
Codex 0 Created 2026-10-03 Updated 2026-10-14
In a parametric statistical model , the likelihood function measures the relative support given by the observed sample for different parameter values. Maximum likelihood estimation supplies a point estimate; the sampling distributions of suitable statistics calibrate significance tests and confidence intervals. A sufficient statistic retains all parameter-dependent information in the likelihood function, so one first seeks a reduction to such a statistic rather than discarding information arbitrarily.
The characteristic conditioning principle in Fisherian conditional inference is to calibrate inference conditional on the observed value of an ancillary statistic. With no nuisance parameter, a statistic is ancillary when its law is the same for every . Its observed value can nevertheless describe the experimental configuration or precision. Conditional inference uses for an informative statistic , with the observed ancillary value. Since the marginal law of does not depend on , conditioning leaves the parameter-dependent likelihood factor unchanged, while potentially changing the appropriate repeated-sampling reference distribution. Conditioning on a continuous means using a regular conditional distribution, rather than dividing by .
With parameters , where is the target and a nuisance parameter, a fully ancillary statistic has a law independent of both. A partial ancillary statistic for may have a law depending on but not on . To eliminate by conditioning, one needs the actual conditional law of the informative statistic to be free of ; partial ancillarity by itself does not guarantee this. For example, if independent counts have Poisson distributions with means and , their total has Poisson distribution with mean , independent of . The first count conditional on has binomial distribution with parameters . Conditioning removes the nuisance intensity exactly. More generally, nuisance-free pivotal quantities or conditional likelihoods can be used where available; there is no universal conditioning device that removes every nuisance parameter.
For a concrete example of non-uniqueness of maximal ancillary statistics, consider one observed pair , withBoth and have Bernoulli distribution with success probability , independent of , so each is an ancillary statistic. Their generated partitions are different and incomparable. Moreover, each is maximal: refining either two-point cell singles out a point whose probability is or , which is parameter-dependent. Their joint statistic is not ancillary, since . Thus there is no common finer ancillary partition that reconciles the two choices. The conditional laws differ as labelled experiments: and , but they condition on different observed information. This illustrates why “condition on a maximal ancillary” need not specify a unique procedure.
For the location-scale family, write , where the are independent with the fixed density . Let . The transformation givesand thereforeFor , almost surely, because an independent sample from a density has probability zero of all observations coinciding. The right side depends only on a sample from the fixed base law, so the entire vector has a law independent of : No finite population mean or variance is needed for this algebraic argument. With one observation the displayed standardization is undefined, so at least two observations are implicitly required.
The pair together with reconstructs the full sample through . Thus conditional inference in a location-scale family useswhere is the observed residual configuration. Equivalently one can use the parameter-free conditional law of given . To see this explicitly, the centered residual subspace has dimension , so its radial coordinate contributes a Jacobian determinant factor proportional to . The conditional density of , is therefore proportional toSetting and gives a conditional density proportional to , with no unknown parameter. The normalizing factor can depend on , but not on or . These statements concern almost every attainable ancillary value.
Past exam of the mathematics course of the University of Cambridge 2008 iii Paper 41 3 Solution by
Codex 0 Created 2026-10-03 Updated 2026-10-14
A Poisson distribution has probability mass , where . It is an exponential family with , mean and variance . The responses are independent, their linear predictors are , and the log link relates these predictors to their means. These are precisely the components of a generalized linear model; the log link is its canonical link function and the dispersion is one.
The log-likelihood and score function areThus a finite maximum-likelihood estimator satisfiesThe Hessian is ; with a full-rank design matrix it is negative definite at finite parameter values, so a finite solution is the unique maximum. A finite maximum need not exist for every possible response/design combination; the displayed equations describe it when it does exist. If the first component of every is one, the first score equation gives .
The saturated model fits each mean as , with the limiting mean zero allowed for a zero count. Subtracting the fitted log-likelihood from the saturated log-likelihood gives the Poisson deviancewhere . With an intercept, the sum of the second terms is zero, but retaining them is necessary without that score equation.
Take low pH as a reference level. For group , the model with interaction haswhere is biomass. The additive model sets every to zero. The interaction model estimates six parameters and the additive model four. In the additive model the log-mean curves are parallel straight lines, while the mean curves are positive exponentials with fixed ratios . In the interaction model the log-mean slopes differ, and the mean ratio is . The following sketches illustrate common versus factor-specific slopes in Poisson regression; their coefficients are illustrative, not estimates from the data.
The additive model is nested in the interaction model with two restrictions. The likelihood-ratio test therefore useswhose p-value is . There is strong evidence that the biomass slope on the log-mean scale depends on pH. The interaction-model residual deviance is close to its residual degrees of freedom, giving no obvious indication of overdispersion from this comparison alone. The supplied deviances contain no slope estimates: they do not establish which pH group has more species, whether each slope is positive or negative, or whether any actual curves cross. Those conclusions require the fitted coefficients and their uncertainty.
Past exam of the mathematics course of the University of Cambridge 2008 iii Paper 41 2 i by
Codex 0 Created 2026-10-03 Updated 2026-10-14
Past exam of the mathematics course of the University of Cambridge 2008 iii Paper 41 1 Solution by
Codex 0 Created 2026-10-03 Updated 2026-10-14
Write . In this normal linear model, the ordinary least squares criterion is . Differentiating with respect to gives the normal equations for linear least squares, . Full column rank of the design matrix makes invertible, and completing the square proves that the unique minimum isSince , its sampling distribution is the multivariate normal distributionLet be the hat matrix. It is symmetric and idempotent with rank , and . The regression residual vector is , soAn orthogonal projection of a Gaussian vector onto the residual space gives . The residual projection and fitted projection have zero cross-covariance, hence are independent because they are jointly normal. In particular, is an unbiased estimator of and is independent of . The maximum-likelihood estimator of instead divides by and is not unbiased.
For the null hypothesis , put . Under that null hypothesis, is standard normal and independent of . Therefore the Student t-test usesA two-sided level- test rejects when . The denominator is the estimated standard error of the regression coefficient.
The fitted linear regression includes an intercept and three covariates:Thus , and the residual degrees of freedom are . The fitted conditional mean isEach slope measures a conditional association with the response while the other two covariates are held fixed. Increasing wind speed by one mile per hour corresponds to a fitted decrease of in the concentration units used. Its standard error is , giving and a two-sided p-value of : evidence of a negative conditional association at the 5% level. The temperature slope is per degree Fahrenheit, with standard error , and p-value . The insolation slope is per langley per day, with standard error , and p-value . Neither of these last two conditional effects is significant at 5%; that does not establish that either true slope is zero. These are associations, not a demonstration of causal effects.
The intercept estimates the mean at zero values of all three covariates. Its value has a large standard error , giving and p-value . The zero-covariate combination may be outside the observed range, so this test may have little practical relevance. Every coefficient test uses a Student t-distribution with degrees of freedom.
The reported residual standard deviation measures unexplained variation in concentration units; the subsequent table gives , consistent with after rounding. The five-number residual summary describes the spread, from to , with median . An intercept forces the residual sum to zero, not the residual median. These five numbers alone cannot establish normal errors, constant variance or independence; residual-versus-fitted and quantile-quantile plots would be more informative, and successive days also warrant checking serial dependence. The coefficient of determination means that 65.33% of the sample response sum of squares about its mean is explained by this fit. Its adjusted version iswhich accounts for the number of estimated regression coefficients.
The stepwise selection by the Akaike information criterion searches between the intercept-only model and the specified model with all three main effects. For these Gaussian fits, the printed Akaike information criterion is, up to a constant common to the candidates,The full model has and AIC . Removing insolation increases the residual sum of squares by but decreases AIC to , so that removal is chosen. Removing temperature or wind instead would give AIC or and is worse. In the next step the candidate models include deletions and restoration of insolation. Keeping wind and temperature has the smallest AIC: removing temperature gives , restoring insolation gives , and removing wind gives . Hence the search stops, givingThe accompanying F-tests compare a larger model with a one-parameter reduction: , where is estimated from the larger model. For example, removing insolation gives , equal to the corresponding squared test statistic. Deletions from the second-stage model use in the denominator, whereas restoring insolation uses the full-model denominator. These tests describe the comparisons; they do not determine the AIC choice. In particular, temperature is retained despite its deletion-test p-value . AIC balances fit and complexity, and a stepwise optimum need not be a global optimum over a richer model class. Ordinary coefficient inference after choosing a model does not account for the selection itself.
For positive responses, the Box–Cox transformation is for , and . Its profile likelihood refits the linear regression and variance at each and includes the transformation Jacobian. The graph peaks at roughly –. The approximate 95% likelihood-ratio confidence interval contains values a little below zero through approximately one, using the cutoff . Thus a logarithmic transformation () and a square-root transformation () are plausible, while leaving the response on its original scale () is at or very near the upper confidence boundary. The broad profile supports considering a mild power transformation; it does not determine a sharply estimated power. The raster does not justify more precise endpoints or a decisive assertion about which side of the boundary lies.
Past exam of the mathematics course of the University of Cambridge 2008 iii Paper 40 4 Solution by
Codex 0 Created 2026-10-03 Updated 2026-10-14
Assume the regime is observed and . With , the wealth dynamics areLet be the value function conditional on the continuous-time Markov chain being in state . Its generator adds to the diffusion generator. The Hamilton-Jacobi-Bellman equation isFor , , the maximizing controls are and . Put . The optimized equation isHere is the utility conjugate.
Choose the normalization ; an additive utility constant shifts every value by that constant divided by . Homogeneity gives with . The regime-switching Merton equations reduce the differential system to one nonlinear algebraic equation per regime:The candidate feedback controls areChanging the regime changes consumption's wealth coefficient, while the risky fraction uses the current regime's risk premium and volatility.
Numerically solve these finite-dimensional equations by a damped Newton method, using to enforce positivity. For , the Jacobian in the variables isMultiply its th column by in the logarithmic variables. A residual-decreasing line search and continuation from , where the positive solution is the scalar Merton coefficient in each well-posed regime, give a practical method whenever that starting problem is finite. More generally use a positive initialization for the coupled system, and check residuals, positivity and conditioning rather than accepting an arbitrary algebraic root.
Verification requires constructing admissible feedback controls, proving the resulting wealth stays nonnegative, and establishing the stochastic-integral and transversality bounds needed at infinite horizon. Apply the Itô formula for semimartingales with jumps to , stopping before wealth leaves a compact interval. The Hamilton-Jacobi-Bellman equation makes accumulated utility plus discounted candidate value a local supermartingale for every admissible policy and a local martingale under the feedback. Justify removal of localization and let the terminal discounted value vanish under the candidate. For , the nonnegative terminal value can be dropped for the upper bound; for it is negative, so a transversality or dual integrability argument is essential. This establishes the global value bound and attainment, rather than just stationarity of the algebraic equations. The boundary is for and the limiting value for .
Past exam of the mathematics course of the University of Cambridge 2008 iii Paper 40 3 Solution by
Codex 0 Created 2026-10-03 Updated 2026-10-14
Use and from Question 1, and take the finite-value case . After retirement the value is . Working for an infinitesimal extra interval while using the retired portfolio and consumption produces a gain . Its zero isThis proves the instantaneous indifference level appearing in the question. For the stated irreversible optimal stopping problem it is generally a lower bound on the retirement boundary, rather than the boundary itself: retaining the opportunity to keep working has value. The following explicit wealth-variable Legendre dual calculation exhibits the correction. In particular, imposing continuity of at retirement would discard that option value; smooth fit requires continuity of .
Assume the usual borrowing-against-income solvency condition: while working, with nonnegative wealth after retirement, and take , , . The question does not specify a tighter borrowing constraint. Let , so and . The working Hamilton-Jacobi-Bellman equation becomes the linear Cauchy-Euler equationThe retired dual is . Let be the negative root ofThe other root exceeds one. Its growing term is excluded by the solvency asymptotic as . Therefore the working dual has the formRetirement corresponds to , where . Value matching and smooth fit giveConsequently the retirement boundary with an income option isUsing the root equation, , which lies strictly between zero and one. Thus for a nonzero risk premium. For example, and give , , andThese parameters give an explicit counterexample to identifying the printed level with the irreversible retirement boundary.
For , find the unique fromThis right side decreases strictly from to , since . Recover the value and controls byFor , retire immediately and use the Merton controls. Along the working strategy, satisfies , and the optimal retirement time is its first hitting time of , equivalently wealth's first hitting time of .
To check the optimal stopping inequalities, write . It satisfies and for . Hence continuation has positive option value there. On the stopping candidate has working residual . The matched dual is continuously differentiable, so the Itô formula introduces no boundary local-time term. Localization, the state-price budget, and the appropriate infinite-horizon transversality justify verification.
If nonnegative financial wealth is imposed during employment as an additional constraint, the working dual also has an upper boundary, and that extra constraint must be included in the free-boundary problem. It cannot be recovered by assuming the printed myopic equality.
Past exam of the mathematics course of the University of Cambridge 2008 iii Paper 40 2 Solution by
Codex 0 Created 2026-10-03 Updated 2026-10-14
In a price-taking equilibrium, market clearing fixes aggregate consumption at the dividend , so and . Each agent takes this aggregate benchmark as given when varying their own consumption. Let be a common state-price density and their budget multipliers. Marginal pricing givesWrite and . ThenFor fixed , the second left side is continuous and strictly decreasing from infinity to zero, so it determines a unique positive number . Every coefficient in that equation is deterministic. ConsequentlyThis is the relative-consumption share equilibrium. The absolute consumption streams remain random through .
Assuming finite fundamental prices, the ex-dividend price of the productive asset follows from conditional pricing:The initial state-price budget constraints determine the multipliers up to a common normalization:Together with the scalar market-clearing equation these characterize the equilibrium. A common scaling of all is offset by the reciprocal scaling of . At least one redundant budget equation can be removed, since the initial fractions sum to one.
For completeness, define and . The consumption-financing wealth is . Its stochastic exposure and that of are both proportional to . Thus holding units of the productive asset, with no bank-account balance, finances this wealth and consumption. Indeed, the holding's deterministic changes and dividend consumption balance because and .
If all and , market clearing gives for a constant , and is constant. The budget equations then implyFor differentiable , the general equilibrium has short rate and risky-asset total expected return ; its volatility is . These follow directly from and the Itô formula. In the identical-agent case, .
Past exam of the mathematics course of the University of Cambridge 2008 iii Paper 4 6 Solution by
Codex 0 Created 2026-10-03 Updated 2026-10-14
A finite-dimensional group algebra has finite representation type of a group algebra when there are only finitely many isomorphism classes of finite-dimensional indecomposable modules. The definition concerns all dimensions, not merely finitely many modules in each fixed dimension.
Let be the block idempotent and choose a maximal Brauer pair : is a block idempotent of with . The inertial index of a block iswhere is the stabilizer of the chosen pair. Changing the pair by conjugation does not alter this number. For cyclic , it is a prime-to- divisor of . Here , as required for the order- subgroup to exist.
Put . Since is characteristic in , , so Green correspondence applies. In fact the nonprojective modules in have a particularly simple correspondence. If , a nontrivial intersection would contain the unique order- subgroup of both cyclic groups, giving and hence , a contradiction. Thus every such intersection is trivial. The error modules in induction are projective. In restriction, an intersection cannot contain a conjugate of a nontrivial subgroup of by an element of : such a conjugate contains , and that would again force . The general Green correspondence therefore pairs all indecomposable modules with nontrivial module vertex contained in a conjugate of . In particular it pairs the nonprojective modules of with the nonprojective modules of its Brauer correspondent in ; the block-compatible form of the correspondence selects . Explicitly,where is projective and the restriction error family is . One should not replace this restriction error term by projectives without an additional argument about .
For the final structure calculation we use the general cyclic-defect structure theorem, which is permitted here as a general fact: the basic algebra of a split block with cyclic defect group is a Brauer tree algebra. Its tree has edges, one simple module per edge, and exceptional multiplicityFor a one-edge tree the basic algebra presentation is . This is the local presentation of the tree algebra, not an assertion that an arbitrary block with one simple module is automatically uniserial.
With the tree has one edge and . Thus the basic algebra of isIt has exactly one simple module, . Its ideals are precisely : for any nonzero element, factor out its lowest power of , after which the remaining factor is a unit. The regular projective therefore has the unique submodule chainwith each consecutive factor one-dimensional and simple. A Morita equivalence from to preserves simple modules, projective covers, submodule lattices and composition length. HenceIn the split setting one may also write . The projective cover has composition factors all isomorphic to , so its vector-space dimension is , not generally .
There is also an elementary reason why the uniserial conclusion follows once the one-simple-module conclusion and finite type are known. If the basic local algebra had , quotienting its radical square and all but two generators would give . The two-dimensional parameter family from Question 3(b) would then give infinitely many indecomposables. Thus is one-dimensional. Nakayama lemma makes generated by one element , so its powers yield a chain and the basic algebra is . The cyclic-defect calculation above identifies . This explains why finite representation type rules out branching in this projective cover.
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!
Intro to OurBigBook
. Source. We have two killer features:
- 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-calculusArticles 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/derivativeVideo 2. OurBigBook Web topics demo. Source. - 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.
- to OurBigBook.com to get awesome multi-user features like topics and likes
- as HTML files to a static website, which you can host yourself for free on many external providers like GitHub Pages, and remain in full control
Figure 2. You can publish local OurBigBook lightweight markup files to either OurBigBook.com or as a static website.Figure 3. Visual Studio Code extension installation.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. - Infinitely deep tables of contents:
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






