In this study we assessed logistic regression and machine learning models to explore their performance in predicting evacuation decisions provide readers with insights into the accuracy of these methods. We tested seven algorithms, including classification tree, Naïve Bayes, K-nearest neighbours, support vector machine, random forest, extreme gradient boosting, artificial neural network. used data collected from 1,807 participants through web-based experiments train calibrate models. The...

This study surveyed citizens (n=1.501) from the Czech Republic, Greece, and Spain to analyse public acceptance of new technologies for preventing terror attacks in urban areas. Our results reveal that threat perception privacy concerns impact on pre-existing attitudes toward surveillance technologies. We found 25% participants trusted proposed while 50% saw them as effective but invasive. Results also sociodemographic factors significantly shape including age, gender, education, political...

Abstract Counter-terrorism measures and preparedness play a critical role in securing mass gatherings, soft targets, infrastructures within urban environments. This paper introduces comprehensive Decision Support System developed as part of the S4AllCitites project, designed to seamlessly integrate with existing legacy systems Smart Cities. The system encompasses pedestrian vehicular evacuation, incorporating predictive models anticipate progression incendiary shooting attacks, alongside...

The stay/go decisions of people involved in fire incidents are crucial for safety. However, the factors that influence these remain unclear. To address this issue, online experiments were conducted to explore responses individuals (n = 1.807) alarms under various conditions. Responses analysed using binary logistic regression. results showed being an enclosed environment and observing others leaving contributed decision evacuate, whereas remaining hindered from evacuating. Additionally,...

Abstract Wildfires are increasing in scale, frequency and longevity, affecting new locations as environmental conditions change. This paper presents a dataset collected during community evacuation drill performed Roxborough Park, Colorado (USA) 2019. is wildland–urban interface including approximately 900 homes. Data concerning several aspects of response were through observations surveys: initial population location, pre-evacuation times, route use, arrival times at the assembly point. used...

Delving into human behaviour during emergencies, this study investigates the evacuation decisions made by 1.807 individuals across fire alarm, explosion, and shooting scenarios. Through a series of 18 trials, encompassing diverse environmental social influences, participants faced decision to evacuate or stay. The impact 20 influential factors on emergency decision-making was explored using logistic regression analyses. Across all threat scenarios, significant majority opted stay rather than...

Predicting and understanding mass evacuations are important factors in disaster management response. Current modelling approaches useful for planning but lack of real-time capabilities to help informed decisions as the event evolves. To address this challenge, a Evacuation Management System (EMS) is proposed here, following stochastic approach combining classical models low complexity high reliability. The EMS computes optimal assembly points shelters related network evacuation routes using...

Population sizing is essential in sociology and various other real-life applications. Gigapixel cameras can provide high-resolution images of an entire population many cases. However, exhaustive manual counting tedious, slow, difficult to verify, whereas current computer vision methods are biased known fail for large populations. A design unbiased method based on geometric sampling has recently been proposed. It typically requires only between 50 100 counts achieve relative standard errors...

Nowadays, a major safety challenge in rail transport is to manage the incidents and emergencies most efficient possible way. The current contingency plans tend be based on static procedures not taking into account how real-time conditions affect them. Consequently, decision-making process may well suffer delays possibility of occurrence for human mistakes could raise since required measures are expected carried out under important pressure. In this study, focused commuter trains, railway...

Population size estimation is relevant to social and ecological sciences. Exhaustive manual counting, the density method automated computer vision are some of methods that currently used. Some these may work in concrete cases but they do not provide a fast, efficient unbiased general. Recently, CountEm method, based on systematic sampling with grid quadrats, was proposed. It offers an can be applied any population. However, choosing suitable parameters sometimes cumbersome. Here we define...

The nucleator is a design unbiased method of local stereology for estimating the volume bounded object. only information required lies in intersection object with an isotropic random ray emanating from fixed point (called pivotal point) associated For instance, neuron can be estimated its nucleolus. extensively used biosciences because it efficient and easy to apply. estimator variance reduced by increasing number rays. In earlier paper systematic sampling was proposed, theoretical...

Population size estimation is essential in ecology and conservation studies. Aerial photography can facilitate this laborious task with high resolution images. However, images thousands of individuals exhaustive manual counting tedious, slow difficult to verify. Computer vision software may work under some particular conditions but they are generally biased known fail several situations. The CountEm a simple alternative based on geometric sampling. It provides fast unbiased for all sorts...

Abstract In this paper we address the use of Neural Networks (NNs) for assessment quality and hence safety several Random Number Generators (RNGs), focusing both on vulnerability classical Pseudo (PRNGs), such as Linear Congruential (LCGs) RC4 algorithm, extending our analysis to non-conventional data sources, Quantum (QRNGs) based Vertical-Cavity Surface-Emitting Laser (VCSEL). Among results found, have classified generators capability NN distinguish between RNG a Golden Standard (GSRNG)....

Abstract Evacuation modelling is continuously open to new scenarios and applications. This study examines the possibility simulate predict evacuation of workers aboard vessels in dockyards. First, we provide data for a better understanding quantification workers’ performance. Second, use an existing model (STEPS) apply validation protocol from observed 150 during unannounced drill Ro-Pax ferry repair period dry dock. Despite uncertainty initial conditions configure scenario, accurately...

Abstract Many of the methods used for estimating population size from ecological surveys have limitations on precision, cost, and/or applicability. The CountEm method was proposed recently number individuals in large groups single images. It is simple and efficient, can be applied to any species. Here we present a case study by applying real survey with 278 images Greater Snow Geese ( Anser caerulescens atlanticus ) Common Eiders Somateria mollissima flocks taken fixed‐wing aircraft Eastern...

Abstract The Eastern Canada (ECA) Flocks data set consists of manually annotated images from the Common Eider (COEI, Somateria mollissima ) Winter Survey and Greater Snow Geese (GSGO, Anser caerulescens atlanticus Spring Survey. were taken in eastern using fixed‐wing aircraft with ImageJ’s Cell counter plugins. We selected ECA order to test precision CountEm flock size estimation method. includes 179 COEI 99 GSGO single images. cut each image a rectangle that excluded large parts no birds....

In this study we conducted a comparison between logistic regression (LR) and machine learning (ML) models to determine the effectiveness of each method in predicting evacuation decisions. We tested seven ML algorithms, including classification tree (CART), Naïve Bayes (NB), K-nearest neighbors (KNN), support vector (SVM), random forest (RF), extreme gradient boosting (XGBoost), artificial neural network (ANN). used data collected from 1.807 participants through web-based experiments train...

The isotropic Cavalieri design is based on a isotropically oriented set of parallel systematic sections constant distance apart. Its advantage over the ordinary twofold - first, besides volume it allows unbiased estimation surface area, and second, error variance predictor for estimator much simpler, involving only area object, between sections. In an earlier paper, two hemispheres rat brain were arranged perpendicular to each other before sectioning, aiming at reducing with respect...

The nucleator is a method to estimate the volume of particle, i.e. compact subset ℝ3, which widely used in Stereology. It based on geometric sampling and known be unbiased. However, prediction variance this estimator non-trivial depends underlying scheme.We propose well established tools from quasi-Monte Carlo integration address problem. In particular, we show how theory reproducing kernel Hilbert spaces can for estimators idea reduced using lattice (or lattice-like) points. We illustrate...