Searcharxiv⌕ Search

arXiv subjects

Carlos M. Hernandez-Suarez

Publications and source records attributed to Carlos M. Hernandez-Suarez.

2 recordsLinked to original sources

Why pyrotechnics markets keep killing:a simple geometric argument for redesign

Fires and explosions in pyrotechnics retail markets recur worldwide with predictable regularity, killing dozens to hundreds of people in single events. This paper argues that the global topology of the market is the dominant determinant of mortality, acting through two independent geometric channels. The first, propagation, concerns ballistic dispersal of ignited articles: the probability that fire spreads between blocks scales with the spatial density of blocks within the dispersal range. The second, evacuation, concerns the distance an occupant must traverse to reach the perimeter, which is set by the global geometry of the market footprint, not by any stall-level parameter. Because mortality risk grows approximately exponentially in evacuation time, topology amplifies modest differences in egress distance into large differences in casualties. Current standards in the United States, the European Union, and Mexico prescribe local parameters such as aisle width and stall separation, but leave the global topology of the market unregulated. We argue that topology should be a regulable design variable, and propose a market geometry that simultaneously slows propagation and shortens evacuation, derived from contact-process models of seed dispersal in spatial ecology.

stat.AP↗

Statistical comparison of Hidden Markov Models via Fragment Analysis

Standard practice in Hidden Markov Model (HMM) selection favors the candidate with the highest full-sequence likelihood, although this is equivalent to making a decision based on a single realization. We introduce a \emph{fragment-based} framework that redefines model selection as a formal statistical comparison. For an unknown true model $\mathrm{HMM}_0$ and a candidate $\mathrm{HMM}_j$, let $μ_j(r)$ denote the probability that $\mathrm{HMM}_j$ and $\mathrm{HMM}_0$ generate the same sequence of length~$r$. We show that if $\mathrm{HMM}_i$ is closer to $\mathrm{HMM}_0$ than $\mathrm{HMM}_j$, there exists a threshold $r^{*}$ -- often small -- such that $μ_i(r)>μ_j(r)$ for all $r\geq r^{*}$. Sampling $k$ independent fragments yields unbiased estimators $\hatμ_j(r)$ whose differences are asymptotically normal, enabling a straightforward $Z$-test for the hypothesis $H_0\!:\,μ_i(r)=μ_j(r)$. By evaluating only short subsequences, the procedure circumvents full-sequence likelihood computation and provides valid $p$-values for model comparison.

stat.ME↗