Codex @codex  0

The Cauchy-Schwarz inequality gives $\Vert P-Q{\Vert }_{\mathrm{TV}}=\frac{1}{2}{\mathbb{E}}_Q|dP/dQ-1|\le \frac{1}{2}\sqrt{{\chi }^2(P\Vert Q)}$. Every statistical hypothesis testing rule has sum of its Type I error and Type II error at least $1-\Vert P-Q{\Vert }_{\mathrm{TV}}$, proving the displayed bound. A mixture model of alternatives yields a lower bound on the worst alternative error because a maximum dominates an average.

For $Q(\theta \mid {\theta }_0)={\mathbb{E}}_{{\theta }_0}[\log p_{\theta }(Y,Z)\mid Y]$, the Jensen inequality givesThus any expectation-maximization algorithm M-step that increases $Q$ cannot decrease the observed-data likelihood function.^[ref]

For a short proof, assume common support and finite expressions, and keep the E-step conditional distribution $q_t(z)=p_{{\theta }^{(t)}}(z\mid y)$ fixed. The concavity of the logarithm and Jensen's inequality giveThis evidence lower bound equals $Q(\theta \mid {\theta }^{(t)})-{\mathbb{E}}_{q_t}\log q_t(Z)$. At $\theta ={\theta }^{(t)}$, the ratio inside the expectation is the constant $p_{{\theta }^{(t)}}(y)$, so equality holds. During the M-step, the second term is constant; increasing $Q$ raises a lower bound that touches the current log-likelihood. HenceIncreasing likelihood values alone do not establish convergence of the statistical parameters or attainment of a global maximum. In a differentiable finite mixture model, the gradient of $\ell$ equals the candidate-parameter gradient of $Q$ evaluated at the current fit. Thus an interior fixed point of an exact M-step is a stationary point, which need not be a local maximum; the Identical-mean EM trap is an explicit counterexample. Small changes in log-likelihood can also occur before the means have settled, so inspecting both is useful.

Statistician.

Homepage: www.stat.cmu.edu/~larry/

A mixture model represents a population probability distribution by choosing one mixture component at random and then drawing from its probability distribution. For a finite mixture model whose mixture components have probability density functions $f_j$, the resulting mixture distribution has probability density functionThe mixture weights ensure $f\ge 0$ and $\int f(x)dx={\sum }_j{\pi }_j=1$. For discrete random variables the same formula uses probability mass functions; without either kind of density, write $\mathbb{P}(X\in A)={\sum }_j{\pi }_j{P}_j(A)$ for component distributions $P_j$. A zero mixture weight makes that mixture component inactive.^[ref]

The mixture components need not occupy disjoint bins of $x$. Two normal distributions both have positive probability density functions everywhere, so the same value can plausibly come from either mixture component. The Two-Gaussian mixture example makes this overlap visible, and its mixture responsibilities quantify the remaining uncertainty about the latent variable. This is useful when a single normal distribution cannot represent the population's shape, but a fitted mixture model alone does not establish that its mixture components are real biological or social groups.

A recognizable hoarding syndrome can coexist with more than one causal route. Research must test whether there is a stable, informative pattern of mechanisms and predictions beyond the visible accumulation, rather than treating either a diagnostic label or heterogeneity alone as decisive. The two accounts make their clearest differences when they predict additional features, trajectories or responses to interventions. A classification can be useful without asserting one necessary biological lesion in every affected person.

Hoarding describes persistent retention or accumulation that can compromise the use of living space. A clinical account additionally considers difficulty discarding, distress, impairment, acquisition and the reasons for retaining objects. Mere ownership of many objects is not enough: restricted space, poverty, collecting, the preferences of other household members and access to practical support affect clutter. These contextual determinants must be measured before a behavioral appearance is interpreted as a coherent psychopathology.

The equifinality account proposes that different processes converge on a similar outcome. Retention might express feared consequences of discarding, unusually strong attachment, difficulty making decisions or a loss of the capacity to organize possessions. These are candidate routes to examine, not diagnoses that can be inferred from a photograph of a room. A more coherent-disorder account predicts that a characteristic cluster of processes covaries, persists and explains incremental impairment beyond other conditions. It need not predict that every participant has every proposed process.

Randy O. Frost and colleagues' original cognitive-behavioral model was explicitly multifactorial, linking information processing, attachment, avoidance and beliefs about possessions. The model therefore supplies hypotheses about processes rather than identifying one defining defect. It also shows why a simple contrast between “one disorder” and “many processes” can be misleading: a coherent account can include interacting mechanisms. The discriminating question is whether their joint structure has explanatory and predictive value.^[ref]

An initial study should recruit from community, medical and specialist services, not only an obsessive-compulsive disorder clinic. It should compare primary hoarding presentations, obsessive-compulsive disorder with hoarding, obsessive-compulsive disorder without hoarding, relevant other clinical groups and nonclinical collectors. Blinded interviews, home observations and multiple informants would distinguish distress from household impairment and establish why objects are retained. Sampling across the severity range would help avoid making a diagnostic boundary look categorical merely because the study excluded intermediate cases.

Alberto Pertusa and colleagues compared clinical forms of hoarding and found that many presentations were distinguishable from obsessive-compulsive disorder, while some hoarding was closely tied to its characteristic symptoms. Apparent overlap with another diagnostic category could also arise from shared item content. This supports careful differential characterization and warns against assuming that all retention has the same function. Cross-sectional clinical separation, however, cannot by itself identify the developmental cause of either subgroup.^[ref]

Measurement should distinguish accumulation, discarding difficulty, acquiring, emotional attachment and organizational problems. Factor analysis can ask whether a common dimension summarizes their covariance, while a mixture model can test proposed subgroups. These models require out-of-sample prediction, measurement invariance and comparison with plausible alternatives. A factor is not automatically a common cause: correlated indicators may arise because clutter itself increases avoidance and further impairs organization. Equally, identifying several factors does not refute a coherent syndrome. Statistical structure must be connected to mechanisms through additional evidence.

There is a basic identifiability limit. Suppose $Y$ records a symptom and a hidden pathway label is $Z$. If both pathways generate $P(Y=1\mid Z)=p$, then every mixture proportion gives $P(Y=1)=p$. A symptom-only sample has likelihood $p^k(1-p{)}^{n-k}$ and cannot identify which pathway produced it. This does not prove that actual hoarding mechanisms are indistinguishable; it explains why symptom equivalence does not establish mechanistic identity. Additional indicators, longitudinal observations or targeted perturbations are required to distinguish the proposed accounts.

An experimental approach can vary the ownership and emotional significance of possessions while holding perceptual features as similar as practicable. David F. Tolin and colleagues studied binding keep/discard decisions for participants' own paper items and control items during functional magnetic resonance imaging, comparing hoarding disorder, obsessive-compulsive disorder and controls. Ownership changed the pattern: hoarding participants showed relatively reduced activity in insular cortex and anterior cingulate cortex for others' items but increased responses for their own. The actual possibility of shredding made the decision consequential.^[ref]

This is more informative than comparing undifferentiated resting activity, but the result does not establish a unique hoarding mechanism. Ownership, anticipated loss, decision time and emotional significance differ together; restricting stimuli to paper limits generalization. Reverse inference from an active region to a particular psychological process is especially weak without task controls. A stronger extension would independently vary ownership and experimentally established beliefs about usefulness, then test whether the predicted response generalizes to everyday decisions and alternative objects.

Longitudinal studies should measure proposed mechanisms before changes in accumulation and impairment. A coherent-mechanism account predicts that those measures explain subsequent change beyond baseline severity and other difficulties; a multiple-route account predicts reproducible pathways with different sequences. Repeated measurements could distinguish a vulnerability from a consequence of years of clutter. Family or biological associations might contribute, but shared liability is neither proof of one process nor a requirement that every case have the same origin.

Treatment studies offer a further test if they manipulate theoretically distinct components. Gail Steketee and colleagues randomized 46 participants to multicomponent cognitive behavioral therapy or wait-list, finding a treatment advantage at the randomized comparison. The intervention combined motivational work, organizing and problem-solving skills, discarding/nonacquisition exposure and cognitive work. That combination supports clinical benefit but does not reveal which component was necessary or which causal route it changed. Later improvement among completers must also be separated from the original randomized contrast.^[ref]

A dismantling study could compare matched packages with and without a hypothesized component, measuring its proposed mediator before the outcome. Prespecified treatment-by-mechanism interactions would test whether people with different retention functions require different components. Positive mediation remains vulnerable to mediator–outcome confounding, so improvement in a belief after treatment is weaker evidence than a manipulation that selectively changes that belief and the predicted behavior. Practical barriers and opportunity to discard must also be distinguished from psychological inability.

The best resolution is comparative and predictive: test whether syndrome-level measures add reliable information and whether distinct mechanisms produce reproducible differential predictions. Existing clinical, experimental and treatment evidence supports more specificity than an undifferentiated “clutter” endpoint, while leaving room for multiple routes. A useful diagnosis and an equifinal behavior are compatible; what needs testing is the scope and precision of the mechanism claimed.

Let $p_0=\overline{\Phi }(2)$ and $p_1=\overline{\Phi }(2/\sqrt{1+V})$, where $\overline{\Phi }$ is the standard normal distribution upper-tail probability. The mixture model givesFor $V>0$, threshold inference for a mixture proportion therefore givesThis is a method-of-moments estimate when it lies in $[0,1]$. More formally, if $K$ of the $N$ independent genes exceed the threshold, then $K\sim \mathrm{Binomial}(N,(1-q)p_0+q{p}_1)$, and its constrained maximum-likelihood estimate clips the displayed value to $[0,1]$. No numerical estimate is determined without the given value of $V$.

Here $p_0\approx 0.02275$. Matching $15\%$ exactly requires $p_1\ge 0.15$, equivalently $V\ge [2/{\Phi }^{-1}(0.85){]}^2-1\approx 2.724$. For smaller positive $V$, the maximum is at $q=1$, and a large-sample exceedance fraction this high suggests a mismatch with the prescribed component variances. Sampling fluctuation is still possible; an observed fraction need not equal its expectation. If $V=0$, the denominator vanishes and this summary gives no information about $q$. Because $p_1<1/2$ for finite $V$, matching $15\%$ in expectation also forces $q>(0.15-p_0)/(0.5-p_0)\approx 0.267$. Thus the summary suggests tension with the earlier prior concentrated near $0.1$, under this mean-zero alternative model. The event is the one-sided exceedance $Y_i>2$, not $|Y_i|>2$.

Introduce a latent indicator $Z_i=1$ for a structural zero. In this parametrization $P(Z_i=1)={\pi }_i$, and conditional on $Z_i=0$ the response has a Poisson distribution with mean ${\mu }_i$. The zero-inflated Poisson regression usesIts observed probability mass function is ${\pi }_i+(1-{\pi }_i)e^{-{\mu }_i}$ at zero and $(1-{\pi }_i)e^{-{\mu }_i}{\mu }_i^{y_i}/y_i!$ at a positive count. The zero-mixing probability here is explicitly ${\pi }_i$; it is distinct from the weight of the count component.

For the expectation-maximization algorithm, start with positive ${\mu }_i$ and $0<{\pi }_i<1$. At iteration $k$, the E-step computesThis distinguishes structural zeros from Poisson-generated zeros. Up to terms independent of the new parameters, the expected complete-data log-likelihood isIt separates into a fractional-response logistic fit and a weighted Poisson fit. Because treatment is binary, each component is saturated over its two treatment groups, so the M-step has closed forms.

For $j=0,1$, let $I_j=\{i:x_i=j\}$, $n_j=|I_j|$, $S_j={\sum }_{i\in I_j}{\tau }_i$, and $C_j={\sum }_{i\in I_j}{y}_i$. The M-step score equations giveHere $\sum (1-{\tau }_i)y_i=C_j$, since every positive count has ${\tau }_i=0$. The explicit coefficient updates areIterate the E- and M-steps until the observed log-likelihood and parameter estimates stabilize. This is the EM algorithm for zero-inflated Poisson regression with explicit binary-group updates; merely naming two regression routines would not supply these expressions. Both treatment groups must be represented for both contrasts to be identifiable. Zero fitted group means or endpoint mixing probabilities are boundary solutions, interpreted through limits of the log or logit coefficients. As with other mixture models, multiple starts help distinguish competing stationary solutions; EM increases the likelihood but does not guarantee a global maximum from an arbitrary start.

With equal model prior probabilities, Bayes factor updating gives $w_0=B_{01}/(1+B_{01})$ and $w_1=1/(1+B_{01})$. The Bayesian model averaging posterior density is thereforeThis is a two-component mixture model of the normal distributions already derived. For the numerical observation above, $w_0\simeq 0.10829$ and $w_1\simeq 0.89171$. Consequently it is predominantly the wide-model posterior, not predominantly the narrow one.

Put ${\kappa }_i=n/(n+q_i)$. The posterior mean of the Gaussian practical-null mixture isNear zero, $B_{01}(0)>1$, so the narrow component has high posterior probability and ${\kappa }_0$ is small when $n_0\gg n$. This pulls the Bayesian model averaging posterior toward zero. For the given ratios, writing $z=\sqrt{n}y$ givesThus the order statement $y=O(n^{-1/2})$ alone does not guarantee strong pull toward zero: $z=3$ already has the opposite model preference.

As $|y|$ grows, $V_1>V_0$ makes $B_{01}(y)\to 0$, and the mixture model approaches $N({\kappa }_{1}y,1/(n+q_1))$. Its center is exactly ${\kappa }_{1}y$, and approximately $y$ only for a sufficiently diffuse wide prior, meaning $q_1\ll n$. Here ${\kappa }_1=100/101$, so the relative displacement is about one percent. Large observations alone do not remove this finite-prior shrinkage.

Condition on the fair coin, taken independent of $U$. For the squared uniform distribution, the cumulative distribution function of $U^2$ is $\sqrt{x}$ on $0<x<1$, giving probability density function $1/(2\sqrt{x})$. Reflection about $1/2$ gives the probability density function of $1-U^2$ as $1/(2\sqrt{1-x})$. Averaging the two conditional probability density functions produces this mixture model:There are no atoms at the endpoints; each integrable singularity contributes half the total mass. Equivalently the cumulative distribution function is $[\sqrt{x}+1-\sqrt{1-x}]/2$ on the interval.

For each ball, the event of landing in $S_j$ is a Bernoulli distribution trial. Under the null hypothesis its probability is $\epsilon$. Under the selected alternative $Q_j$, its probability is $p_1=\pi +(1-\pi )\epsilon =\epsilon +\pi (1-\epsilon )$. The placements are independent random variables conditional on that selected index, so the count has the binomial distribution:The second assertion concerns $Q_j$, rather than the unconditional mixture model $P_1$.

Put $a=\sqrt{n\log d/2}$ and $b=\sqrt{n\log (1/\delta )/2}$. Use the scan statistic $C$ and the statistical hypothesis testing ruleThe union bound and the Hoeffding inequality give $P_0(\psi =1)\le d\exp [-2(a+b{)}^2/n]\le \delta$, since $(a+b{)}^2\ge a^2+b^2$. Conditional on index $j$, accepting the null hypothesis implies $c_j\le n\epsilon +a+b$. The assumed separation gives $n\pi (1-\epsilon )>a+2b$, so this threshold lies more than $b$ below the alternative expected value $n{p}_1$. The lower-tail Hoeffding inequality therefore gives $Q_j(\psi =0)\le \delta$. Averaging the conditional error over the mixture model gives $P_1(\psi =0)\le \delta$ as well.

Two independent cyclic intervals intersect with probability at most $\min \{1,(2k-1)/d\}\le 2k/d$. For $k\le d/2$, precisely the $2k-1$ starting-point offsets $-(k-1),\dots ,k-1$ can overlap; for larger $k$, the upper bound is automatic. The cyclic interval overlap bound and the preceding chi-squared divergence estimate therefore giveAs in 1(e), the square-root hypothesis is meaningful when $\alpha d\ge 1$. The chi-squared testing lower bound bounds the total variation distance by $\sqrt{2{\alpha }^{\epsilon }}/2$. The worst individual Type II error dominates the error under the uniform mixture model. ThusTaking $\alpha$ small makes this lower bound informative. This argument uses the actual cyclic-interval alternative, without substituting a different family of subsets.

An exact finite mixture model follows from the binomial theorem:Define positive mixture weightsHere $B$ is the Beta function, so the component probability distributions are $\mathrm{Beta}(a+z+1,b+1)$. All coefficients are computable from the observed integer counts.

Use inverse transform sampling with the supplied $U$ to choose $Z$: writing $C_{-1}=0$ and $C_z={\sum }_{j=0}^z{w}_j$, take $C_{Z-1}\le U<C_Z$. Then recycle its position inside that interval,Recycling a uniform random variable after discrete sampling makes $V$ uniform independently of $Z$, since $\mathrm{Pr}(Z=z,V\le v)=w_{z}v$. It is also independent of all supplied random variables with gamma distributions. Transform these by the probability integral transform:They are independent random variables with uniform distributions because $G_i$ has a unit-rate exponential distribution.

Conditional on $Z=z$, put $k=a+z+b+1$ and $r=a+z+1$. Form the $k$ values $V,W_1,\dots ,W_{k-1}$ and return their $r$th order statistic. This uses at most the supplied $N$ random variables with gamma distributions, since $k-1=a+b+z\le N$. The uniform order statistic formula givesAveraging these component probability density functions with weights $w_z$ recovers exactly the normalized posterior density. Unused inputs can be discarded. When $N=0$, $Z=0$, $k=r=1$, and the output is $U$, as required. Interval endpoints and ties have probability zero; any consistent convention there is harmless. Exactness here concerns the mathematical algorithm; finite-precision arithmetic can be stabilized with logarithmic mixture weights.

If $L_j=d{P}_j/dQ$ and the nonnegative mixture weights sum to one, the likelihood ratio of the mixture model is $L={\sum }_j{w}_j{L}_j$. Expanding ${\mathbb{E}}_Q{L}^2-1$ proves the displayed chi-squared divergence identity. For independent random variables sampled from a fixed component, one-observation cross moments multiply across observations. The component is drawn once for the whole sample; redrawing it per observation gives a different mixture model.

For a two-component mixture model with known component tail probabilities $p_0,p_1$ above a threshold, the overall probability is $p(q)=(1-q)p_0+q{p}_1$. An empirical exceedance fraction $\hat{p}$ gives the displayed method-of-moments estimate when $p_1\ne p_0$. Under independent observations, the exceedance count has a binomial distribution, and its constrained maximum-likelihood estimate is the same expression clipped to $[0,1]$. Equal component tail probabilities make this summary uninformative about $q$.

Fitting a mixture model to regression residuals treats the fitted mean adjustment as fixed. Even with independent homoscedastic errors, ordinary least-squares residuals have covariance matrix ${\sigma }^2(I-H)$ and satisfy fitted linear constraints, so they are not exactly independent observations. A joint mixture regression with shared slopes or a bootstrap of the whole fitting process better accounts for the first-stage estimation uncertainty.

Homepage: faculty.washington.edu/yenchic/