Isaac Esteban Camargo Freile

Hyperspectral imaging is a well-established remote sensing technique for material detection and classification, relying on hundreds of reflectance bands to exploit the spectral signatures of surface materials. Although hyperspectral imagery has demonstrated excellent capabilities for material identification, its operational implementation may be constrained in some applications due to data volume and processing requirements. Consequently, there is growing interest in evaluating the capability of lower-dimensional multispectral imagery for material detection tasks. In this sense, this article proposes as its contribution the comparative evaluation of machine learning models and neural networks for asbestos–cement detection on VNIR imagery. For the development of this research, the CRISP-DM methodology was adapted into four phases: P1. Business and data understanding; P2. Data preparation; P3. Modelling and evaluation; P4. Model deployment. At the results level, three datasets with different numbers of bands were constructed, which were structured by adding to the original dataset an additional layer with the NDVI and two additional layers with the PCA components of the original image. Across the three datasets, four machine learning models and one neural network model were tuned and evaluated, yielding as a result that in all three datasets the KNN and neural network models achieved the best performance. Likewise, it was found that the detection capability of the models improved with the inclusion of the additional bands. The proposed approach serves as a reference to be extrapolated by research centres and universities for the detection of asbestos and other materials in VNIR images, with a view toward integration into resource-constrained systems and specifically into environmental monitoring systems.

Dengue fever poses an ongoing challenge to global and Colombian public health. Although surveillance microdata are widely available, there remains a gap in converting them into actionable epidemiological intelligence. This study presents a reproducible dataset and analytical resource for transforming routine records into a spatiotemporal framework for territorial risk stratification. To this end, 15 years (2010–2024) of anonymized records from the SIVIGILA in Colombia’s Caribbean region were consolidated, covering 303,801 cases across 197 municipalities. The microdata were aggregated into municipality–year analytical units using seven indicators of magnitude, severity, demographics, and surveillance performance. Subsequently, an unsupervised learning model (K-means) was validated using the Elbow and Silhouette methods. The algorithm consistently identified four heterogeneous epidemiological profiles: high-transmission urban settings, dispersed rural risk municipalities, territories with a pediatric predominance, and clusters of high clinical severity with elevated hospitalization and case fatality rates. In conclusion, this dataset and its methodological framework transform static historical information into an operational tool that facilitates strategic surveillance, the development of interactive dashboards, and the territorial prioritization of evidence-based public health interventions.

Intimate partner violence (IPV) includes assaults that risk a woman’s bodily integrity. Intimate partners commit IPV, people with whom the victim shares (or shared) a close personal or sexual relationship. This phenomenon has a great global and national impact. Thus, it is necessary to establish trends of the risk of physical violence to women by their current or former partner in each department of Colombia and its relationship with sociodemographic and health characteristics. This study uses an ecological approach at the departmental level, with victims of intimate partner violence treated at the National Institute of Legal Medicine and Forensic Sciences (INMLyCF). Potential factors were identified through Bayesian factor analysis and were included in the model to estimate risk. The findings show that the Casanare department had the highest risk of producing victims (SMR: 2.545). In departments where the educational level of women is at or below primary school, there is a high-risk β = 0.343 (0.285, 0.397) of them being assaulted. For the departments in which the employment of women is in sales and services or office workers, the associated factor presents a higher risk β = 0.361 (0.201, 0.485), as in the risk related to affiliation with the social security system β = 0.338 (0.246, 0.498), as well as sexual and reproductive life β = 0.143 (0.003, 0.322). The following categories were associated with physical gender violence: no education and low participation in making purchases at home β = 0.106 (0.049, 0.199), low participation in decisions about their health, and visits to family and friends β = 0.240 (0.170, 0.299). Therefore, public health programs should strengthen women’s empowerment in household decisions and increase their educational level to reduce this incidence.