Abstract: One of the main public health problems is child malnutrition, since it negatively affects the individual throughout his life, limits the development of society and makes it difficult to eradicate poverty. The first objective of this research is to apply data mining techniques for preprocessing, cleaning, reduction and transformation to a data lake that has allowed analyzing anemia in children under 5 years of age, the second objective is to apply Machine Learning algorithms to obtain the best model to predict anemia in children under 5 years of age. The data set was extracted from the open data platform of the government of Peru that corresponds to South Lima, North Lima, East Lima, Central Lima and rural Lima, which collected a total of 138,369 instances and 36 variables of which 30 are categorical and 6 numeric, being an unbalanced data set. In order to obtain the best predictor variables, the Anova F-test and Chi Square filters were used, and it was possible to reduce them to 10 variables, cases were also carried out without considering one of the filters and both filters.To find the best prediction model, the algorithms have been tested: decision tree, logistic regression, K nearest neighbors, random forest and naive bayes. As a result, we show that the best algorithm to predict anemia in children under 5 years of age is the Naive Bayes algorithm with the highest recall of 74%, precision of 43% and accuracy of 70%.

Resumen: Uno de los principales problemas de salud pública es la desnutrición infantil, ya que afecta negativamente al individuo a lo largo de su vida, limita el desarrollo de la sociedad y dificulta la erradicación de la pobreza. El objetivo de esta investigación es aplicar técnicas de minería de datos para el preprocesamiento, limpieza, reducción y transformación a un lago de datos que ha permitido analizar la anemia en niños menores de 5 años, asimismo, se ha aplicado algoritmos de Machine Learning para obtener el mejor modelo que permita predecir la anemia en niños menores de 5 años. El conjunto de datos fue extraído de la plataforma datos abiertos del gobierno del Perú que corresponde a Lima sur, Lima Norte, Lima Este, Lima Centro y Lima rural que se juntaron en total 138369 instancias y 36 variables, de la cuales 30 son categóricas y 6 numéricas, siendo un conjunto de datos desequilibrado. Para poder obtener las mejores variables predictoras se utilizó los filtros Anova F-test y Chi Cuadrado y se logró reducir a 10 variables, también se realizó casos sin considerar uno de los filtros y ambos filtros. Para buscar el mejor modelo de predicción se ha probado los algorítmicos: árbol de decisión, regresión logística, K vecinos más cercanos, bosque aleatorio y Naive bayes. Como resultado mostramos que el mejor algoritmo que permite predecir la anemia en niños menores de 5 años es el Naive Bayes con mayor recall de 74 %, precisión de 43% y exactitud del 70 %.

ABSTRACT This report described temporary hyperpigmentation occurring in a cat after use of the FreeStyle® Libre sensor by applying three different adhesive glues. The device was continuously used as an aid for the treatment of diabetes mellitus. Histopathological examination revealed mild focal chronic active dermatitis. While the coat color change occurred due to inflammation and did not affect the device functioning, owners should be alerted to the possibility of this complication.

RESUMO Este relato de caso descreveu a hiperpigmentação temporária que ocorreu em um gato após o uso do sensor FreeStyle® Libre, no qual foram aplicadas três colas adesivas diferentes. O dispositivo foi usado continuamente como auxiliar no tratamento do diabetes mellitus. O exame histopatológico revelou dermatite crônica ativa focal leve. Embora a mudança de cor da pelagem tenha ocorrido devido à inflamação e não tenha afetado o funcionamento do dispositivo, os proprietários devem ser alertados para a possibilidade dessa complicação.

Abstract The aim of this study was to report on the presence of microfilariae of Dirofilaria immitis causing nodular pyogranulomatous dermatitis in a dog in the state of Rio Grande do Norte, northeastern Brazil. A 4-year-old male dachshund dog with lesions in the nostrils and left dorsolateral regions was treated. Tests were requested to aid in making the diagnosis, such as skin cytology, Knott's test, thick smear and histopathology of the lesions. From these, presence of a diffuse pyogranulomatous process was observed and, amidst the cellular material, microfilariae of Dirofilaria spp. A conventional polymerase chain reaction test on tissue samples from the lesions revealed the presence of the species D. immitis. Treatment based on ivermectin (3mg) was administered at a single oral dose of 0.6 mg/kg. In the first seven days there was regression of the lesions, but after 30 days there was recurrence. A new treatment was administered, consisting of 10% imidacloprid + 2.5% moxidectin (4-10 mg/kg), with one application per month for 6 months, and doxycycline (100 mg), 10 mg/kg, 1 tablet, 2 times a day, for 30 days. In conclusion, D. immitis microfilariae caused pyogranulomatous lesions in the subcutaneous tissue of a dog. This had not previously been described in Brazil.

Resumo O objetivo deste trabalho foi relatar a presença de microfilárias de Dirofilaria immitis causando dermatite nodular piogranulomatosa em cão no estado do Rio Grande do Norte, região Nordeste do Brasil. Um cão de 4 anos, da raça Dachshund, foi atendido com lesões nas regiões das narinas e dorsolateral esquerda. Foram realizados exames de citologia cutânea, teste de Knott, gota espessa e histopatologia das lesões, sendo observada a presença de um processo piogranulomatoso difuso e microfilárias de Dirofilaria spp. Na reação em cadeia da polimerase convencional constatou-se a espécie D. immitis nas amostras de tecido das lesões. Foi efetuado um tratamento à base de Ivermectina (3mg) 0,6 mg/kg – 1 aplicação, via oral. Nos sete primeiros dias, houve regressão das lesões, porém, depois de 30 dias, houve recidiva. Foi efetuado um novo tratamento com imidacloprida 10% + moxidectina 2,5% (4-10 mg/kg), 1 aplicação por mês durante 6 meses, e Doxiciclina (100 mg), 10 mg/kg, 1 comprimido, duas vezes ao dia, por 30 dias. Em conclusão, microfilárias de D. immitis causaram lesões piogranulomatosas em tecido subcutâneo de um cão, relato anteriormente não descrito no Brasil.

Abstract Despite the enormous paleobotanical record on different islands of the Antarctic Peninsula, the evidence of insect activity associated with fossilized plants is scarce. Here we report the first evidence of insect-plant interaction from Cretaceous deposits, more precisely from a new locality at the Rip Point area, Nelson Island (Antarctic Peninsula). The macrofossil assemblage includes isolated Nothofagus sp. leaf impressions, a common component of the Antarctic paleoflora. Two hundred leaves were examined, of which 15 showed evidence of insect activity, displaying variations in size, shape, and preservation. Two types of interaction damage, galls and mines, were identified. A single specimen retained a circular scar recognized as galling scar, while meandering tracks were considered mines. These traces of herbivore insect activity, correspond to the oldest known record of this type of interaction of West Antarctica and the oldest record of insect-plant interaction in Nothofagus sp. reported so far.

RESUMEN Se estableció las características y competencias tecnológicas para la educación virtual en docentes de algunas de las escuelas de formación policial en Colombia. Para esto, se realizó un estudio descriptivo con entrevistas semiestructuradas a 109 docentes de las Escuelas de formación en Vélez (Santander) y en Bogotá D. C., donde se desarrollaron procesos de enseñanza y aprendizaje de manera virtual durante el 2020; para fines de triangulación, se entrevistaron 385 estudiantes de las mismas escuelas. Los resultados mostraron que las competencias de los docentes en su mayoría se encuentran en nivel exploratorio y básico, principalmente en cuanto a la creación de materiales y recursos, la comprensión del modelo pedagógico virtual a partir de las tecnologías de la información y la comunicación, y del dominio de los procesos de seguridad a la información para ser trasferidos a los estudiantes. Los hallazgos permitieron demostrar la existencia de una brecha entre las competencias actuales de dominio tecnológico de los docentes de las escuelas de formación y aquellas propias del perfil del docente policial, lo que puede incidir también en las competencias de los graduados y requiere un abordaje urgente a partir del plan de formación profesoral.

RESUMO Esta pesquisa teve como objetivo estabelecer as características e competências tecnológicas para o desenvolvimento de atividades, recursos e estratégias relacionadas à educação virtual, em professores de algumas das escolas de formação policial da Colômbia. Para tanto, foi realizado um estudo descritivo no qual foram aplicadas entrevistas semiestruturadas a 109 professores das escolas de formação de Vélez (Santander) e da cidade de Bogotá, onde os processos de ensino e aprendizagem foram desenvolvidos de forma virtual durante o ano de 2020; para fins de triangulação, 385 alunos das mesmas escolas foram entrevistados e treinados com os professores participantes. Os resultados mostraram que as competências dos professores estão majoritariamente no nível exploratório e básico, principalmente em termos de criação de materiais e recursos, compreensão do modelo pedagógico virtual baseado nas tecnologias da informação e comunicação e domínio dos processos de segurança da informação a serem transferidos para os alunos. Os achados permitiram demonstrar a existência de um distanciamento entre as atuais competências de domínio tecnológico dos professores das escolas de formação e as do perfil do professor policial, o que também pode afetar as competências dos egressos e requer uma abordagem urgente do plano de formação docente.

ABSTRACT This research aimed to establish the characteristics and technological competencies for the development of activities, resources and strategies related to virtual education, in teachers of some of the police training schools in Colombia. For this purpose, a descriptive study was carried out in which semi-structured interviews were applied to 109 teachers from the training schools in Velez (Santander) and Bogota D. C., where teaching and learning processes were developed in a virtual way during the year 2020; for triangulation purposes, 385 students from the same schools were interviewed and trained with the participating teachers. The results showed that the skills of teachers are mostly at the exploratory and basic level, mainly in terms of creating materials and resources, understanding the virtual pedagogical model based on information and communication technologies, and mastering information security processes to be transferred to students. The findings made it possible to demonstrate the existence of a gap between the current technological mastery skills of the teachers of the training schools and those of the profile of the police teacher, which may also affect the skills of the graduates and requires an urgent approach from the teacher training plan.

ABSTRACT This paper addresses the problem of the optimal reconfiguration of medium-voltage distribution networks by proposing a mixed-integer nonlinear programming (MINLP) model. The objective function of this optimization model is the minimization of the total power losses in all the branches of the network, considering active and reactive power balance equations, voltage regulation bounds, and device capabilities, among others. The proposed MINLP formulation works with branch-to-node incidence that allows representing the active and reactive power flow in branches as a function of the real and imaginary components of the voltages and currents. The solution of the MINLP model is reached through the general algebraic modeling system widely known as the GAMS package while presenting it in the form of a tutorial. This software allows implementing the proposed model a in compact way, which is solved via branch and bound methods. Two test feeders comprising 5 and 14 nodes allow demonstrating the fidelity of the proposed MINLP model regarding power loss minimization when compared to that reported in the specialized literature.

RESUMEN Este artículo aborda el problema de la reconfiguración óptima de redes de distribución de media tensión mediante la proposición de un modelo de programación no lineal entera mixta (PNLEM). La función objetivo de este modelo de optimización es la minimización de las pérdidas totales de potencia activa en todas las ramas del sistema, considerando las ecuaciones de balance de potencia activa y reactiva, los límites de regulación de voltaje y la capacidad de los diferentes dispositivos, entre otros. La formulación de PNLEM propuesto trabaja con la matriz de incidencia rama-nodo, la cual permite representar las ecuaciones de flujo de potencia en las ramas como una función de los componentes reales e imaginariios de los voltajes en los nodos y las corrientes en las ramas. La solución del modelo de PNLEM se obtiene a través del sistema de modelado algebraico general ampliamente conocido como el paquete GAMS, presentándolo en forma de turorial. Este software permite implementar el modelo de optimización propuesto en forma compacta, el cual se resuelve mediante métodos de ramificación y corte. Dos alimentadores de prueba compuestos de 5 y 14 nodos permiten demostrar la fidelidad del modelo de PNLEM propuesto en relación con la minimización de pérdidas de potencia cuando se compara con reportes de la literatura especializada.

RESUMO Objetivou-se, com este estudo, conhecer a percepção dos médicos veterinários quanto à monitorização da glicose sanguínea e intersticial em gatos com diabetes mellitus (DM) e/ou cetoacidose diabética, com ênfase no sistema flash de monitoramento da glicose (SFMG) (FreeStyle Libre, Abbott, Brasil). Esta pesquisa foi composta por duas etapas. Ao todo, foram obtidos 516 formulários de resposta. Desses, 480 (93%) foram considerados válidos. No total, 333 (69,4%) veterinários não utilizavam o SFMG, enquanto 147 (30,6%) o utilizavam. O custo do SFMG (116, 78%) foi a maior barreira para a aceitabilidade. Os médicos veterinários que utilizam o dispositivo consideram-no indispensável no monitoramento hospitalar da cetoacidose diabética por ser um facilitador no acompanhamento preciso das aferições. Além disso, o local de preferência para aplicação do sensor é na parede lateral cranial do tórax e é bem tolerável. Monitorar um felino diabético requer comprometimento do tutor e da equipe veterinária e que assegure um manejo amigo do gato.

ABSTRACT This study aimed to understand the perception of veterinarians regarding monitoring blood and interstitial glucose levels in cats with diabetes mellitus and/or diabetic ketoacidosis, with emphasis on the flash glucose monitoring system (FGMS) (FreeStyle Libre, Abbott, Brazil). This research consisted of two stages. In all, 516 response forms were obtained, and of these, 480 (93%) were considered valid. In total, 333 (69.4%) veterinarians did not use the FGMS, while 147 (30.6%) did. The cost of the FGMS (116, 78%) was the greatest deterrent to acceptability. Veterinarians who use the device consider it indispensable in the hospital monitoring of diabetic ketoacidosis and a facilitator in the accurate monitoring of measurements. In addition, the preferred location for application of the sensor is the cranial lateral wall of the chest and it is quite tolerable. Monitoring a diabetic cat requires commitment from the owner and the veterinary team to ensure feline-friendly management.

Resumen El artículo aborda la oposición teórica de dos modelos espaciales similares, pero conceptualmente diferentes, de cooperación transfronteriza en red (a saber, la integración territorial de las EMN y la integración informal de la red) desde el aspecto de la asimetría de poder corporativo de las empresas que participan en tales estructuras de red. Se presentan dos fenómenos externos de asimetría (simetría) de poder dando como ejemplo los casos de incrustación territorial e integración informal de redes en las regiones de América del Norte y Asia Oriental. El estudio presenta varias características que permiten proponer la atribución de sistemas de producción transfronterizos específicos a diferentes tipos de estructuras de red. Esto último se hace mediante una consideración de las características de la red, los efectos de la red y los detalles de la interacción del núcleo y la periferia de la red.

Abstract This article deals with a theoretical opposition of two similar, but conceptually different spatial models of transborder network cooperation (namely MNEs’ territorial embeddedness and informal network integration) from the aspect of corporate power asymmetry of firms participating in such network structures. Two external phenomena of power asymmetry (symmetry) are presented giving as example cases of territorial embeddedness and informal network integration in North America and East Asia regions. The study is resulted in several features allowing the attribution of specific transborder production systems to be proposed to different types of network structures. The latter is done by a consideration of network characteristics, network effects and specifics of network core and periphery interaction.

Abstract A study of macro and microfacies, palynoflora and palynofacies of the non-marine Cerro Negro Formation at President Head Peninsula, Snow Island, northwest of the Antarctic Peninsula, was developed. Two assemblages were recognized: Palynofacies assemblage 1 (P1) at the base of the section with a dominance of fern spores and conifer pollen grains, and facies association consisting of a clastic layer, with the predominance of mudstones; and Palynofacies assemblage 2 (P2) at the top of the section, with remarkable abundance of AOM/Pseudoamorphous particles, associated with facies that includes tuffs. The complete section shows in some levels the presence of freshwater algae and translucent phytoclasts. The integrated data characterizes a fluvial-lacustrine environment, what is reinforced by the occurrence of freshwater algae (Botryococcus) in some levels of P1 and P2. We could verify an increase in volcanic activity towards the top of the section that apparently has played an important role in the collapse of the palynoflora. The occurrence of the spore species Muricingulisporis annulatus, Sotasporites elegans, S. triangularis, Foraminisporis wonthaggiensis, and F. asymmetricus in the Cerro Negro Formation allows the correlation with sections in South America and Australia, suggesting an Aptian age for these deposits.

Abstract This study is focused on the fate of a large volume of mine slurry discharged from the Doce River (DR) to the coastal ocean after the worst environmental disaster in Brazilian which occurred in November 2015. We used Eulerian (ROMS) and Lagrangian (STRiPE) numerical models, as well as satellite remote sensing data, to study the spreading and seafloor accumulation of fine river-borne sediments during the initial six months following the disaster. We show that the regions of intense sediment accumulation were determined by spreading patterns of the surface-advected DR plume. The river discharge rate governed the plume surface area, while its position depended on local wind forcing conditions. The spreading of sediments carried by the DR plume was dominated by southward transport caused by prevailing upwelling-favorable northeasterly winds during the study period. Under high discharge conditions, river-borne sediments were transported over 100 km southward from the DR mouth and reached the outer shelf. In contrast, sediments were arrested near the mouth during drought periods and remained on the inner shelf. As a result, fine river-borne sediments accumulated on the seafloor, mainly in the large shallow shelf area southward from the DR mouth. Conversely, only a small fraction of residue was deposited northward. Thus, the Environmental Protection Area (EPA) of Costa das Algas, located 40 km southward from the DR, potentially exhibited more susceptibility to sediment arrival. On the other hand, their influence on Abrolhos Marine National Park, located 200 km northeastward from the DR mouth, was presumably minimal.

Abstract The Colombian Amazon region is part of the Neotropical rainforest (humid forest biome) covering an area of 483,163 km2 and includes tributaries of both the Amazon and Orinoco River basins. The aquatic ecosystems found there include: rivers and alluvial plains originating in Andean headwaters, on eroded soils of tropical forests in the lowlands, and Guiana Shield formations, comprising a dense fluvial drainage network in the lowlands, with Paleogene/Neogene geological formations (terra firme streams in higher places that don’t usually flood) and Paleozoic (shield streams); and Andean and Guiana Shield streams above 200–250 m a.s.l. We present here an exhaustive compilation of published information, supported by fish collections, consisting of a list of 1104 species distributed in 375 genera, 53 families, and 16 orders. We include occurrence data of these species in each sub-basin. The presence/absence species matrix was analyzed using a dendrogram and non-metric multidimensional scaling (NMDS) analysis to identify patterns of similarity between basins and sub-basins. We evaluated species composition between basins and among the different geological origins using PERMANOVA. The dendrogram shows co-occurrences of 404 species in the two basins. It also shows two clear groupings of the sub-basins of the Amazon (except Guainía-Negro drainages) and those of the Orinoco. Within the Amazon Basin, there are two nodes according to the geological origin: systems of Andean origin and those of the lowlands. The dendrogram results are consistent with the NMDS analysis, which shows a clear grouping according to the connectivity of the basins; the Guainía-Negro is included in the Amazon basin. Species distribution patterns were supported by the PERMANOVA, and differed significantly between basins (F = 4.3, R = 0.26, P = 0.003) and geological origin (F = 3.6, R = 0.23, P = 0.003). The number of species in this study represents almost a fifth of the ichthyofauna of the Neotropics and about a third of that of the Amazon River basin; clearly supporting Colombia’s status among the countries with the greatest diversity of freshwater fish species of the planet. We include here a significant number of new records (75 spp), provide a first approximation of the distribution patterns, and a framework for future biogeographical studies.

Resumo La región de la Amazonía colombiana hace parte del bosque húmero neotropical (bioma de selva húmeda) abarcando un área de 483.163 km2 e incluye afluentes de las cuencas del Amazonas y Orinoco. Los ecosistemas acuáticos encontrados allí incluyen: ríos y llanuras aluviales, originándose en cabeceras andinas, en suelos erosionados de bosques tropicales en tierras bajas y en formaciones de escudos; conformando una densa red fluvial en tierras bajas con formaciones geológicas paleógenas-neógenas (arroyos de terra firme en sitios elevados que usualmente no se inundan) y paleozoicas (arroyos de escudo); y arroyos andinos y del escudo Guayanés por encima de 200–250 m s.n.m. Presentamos aquí una recopilación exhaustiva de información publicada, sustentada por colecciones ictiológicas, consistiendo en una lista de 1104 especies distribuidas en 375 géneros, 53 familias y 16 órdenes. Incluimos datos de estas especies en cada subcuenca. La matriz de presencia/ausencia de especies fue analizada usando un dendrograma y un análisis de escalamiento multidimensional no métrico (NMDS) para identificar patrones de similitud entre cuencas y subcuencas. Se evaluó la composición de especies entre cuencas y entre los diferentes orígenes geológicos usando PERMANOVA. El dendrograma refleja coocurrencia de 404 especies en las dos cuencas. También muestra dos agrupaciones claras de las subcuencas del Amazonas (excepto Guainía-Negro) y las del Orinoco. Dentro de la cuenca amazónica existen dos nodos según el origen geológico: los sistemas de origen andino y los de tierras bajas. Los resultados del dendrograma son consistentes con el análisis NMDS, el cual muestra una clara agrupación según la conectividad de las cuencas; el Guainía-Negro está incluido en la cuenca del Amazonas. Los patrones de distribución de especies fueron respaldados por el PERMANOVA y difirieron significativamente entre cuencas (F = 4.3, R = 0.26, P = 0.003) y origen geológico (F = 3.6, R = 0.23, P = 0.003). El número de especies en este estudio representa casi la quinta parte de la ictiofauna del Neotrópico y alrededor de un tercio de la de la cuenca del río Amazonas; soportando el estatus de Colombia entre los países con mayor diversidad de especies de peces de agua dulce del planeta. Incluimos aquí un número importante de nuevos registros (75 spp), brindamos una aproximación de los patrones de distribución y un marco para futuros estudios biogeográficos.

Resumen Los aceites esenciales son compuestos orgánicos importantes en el sector farmacéutico debido a sus propiedades antimicrobianas, volviéndose una alternativa de solución para determinadas fitopatologías. En el presente estudio se evaluó in vitro la actividad inhibitoria de aceites esenciales de Curcuma longa y Zingiber officinale contra Colletotrichum sp. obtenidas a partir de cultivos de Dioscorea rotundata en Toluviejo (Sucre, Colombia). Concentraciones de 3,000; 5,000; 8,000 y 10,000 mg/L de extracciones de aceite (método de hidrodestilación asistida por microondas) fueron aplicados a cultivos in vitro del hongo. Además, se empleó un testigo absoluto y controles positivo y negativo. Los aceites fueron caracterizados por cromatografía de gases acoplado a espectrometría de masa. La actividad inhibitoria se realizó por medio de la prueba de siembra en superficie (porcentaje de índice de inhibición). Los mayores valores de índice antifungico se observaron a 10,000 mg/L similar al obtenido con el control positivo por benomil. De la caracterización química de los aceites esenciales el metabolito secundario que se presentó con mayor rango fue el monoterpeno 3-Thujeno y el citral. Los aceites esenciales de C. longa y Z. officinale generaron actividad inhibitoria contra el crecimiento de Colletotrichum sp., y por lo tanto pueden ser empleados para reducir la antracnosis.

Abstract Essential oils are organic compounds during the last century have become important in the pharmaceutical sector because of its antimicrobial properties, being an alternative solution to certain plant diseases. In the present study, we evaluate in vitro the inhibitory activity of essential oils of Curcuma longa and Zingiber officinale against Colletotrichum sp. obtained from yam crops (Toluviejo, Sucre, Colombia). Concentrations 3,000; 5,000; 8,000 and 10,000 mg/L of each species of essential oils extractions were applied in vitro on the fungi (microwave assisted hydro-distillation method). In addition, we employed an absolute control, positive control (benomyl 1g/L) and a negative control. The inhibitory activity was performed by seeding the test surface and expressed as percentage of inhibition index. We obtain than the highest percentages of antifungal index were observed at 10000 mg/L, similar to that obtained with the positive control. Of the chemical characterization of essential oils (gas chromatography coupled with mass spectrometry) the secondary metabolite that occurred more range was the monoterpene 3-Thujeno and citral. The essential oils of C. longa and Z. officinale show inhibitory activity against the growth of Colletotrichum sp., and therefore can be employed to reduce the use of chemical compounds to reduce the anthracnose.

Abstract Superconducting Fault Current Limiter (SFCL) modules with distinct arrangements have been investigated, seeking design and performance enhancements. In this context, this work aims to propose a design for resistive SFCL (R-SFCL) modules, using RE-Ba-Cu-O high-temperature superconducting tapes (HTS) wound on an alternative support material. An R-SFCL bench prototype was designed for 400 A and 500 V. Two pieces of 2G tapes, arranged in a bifilar antiparallel configuration were wound on Acrylonitrile Butadiene Styrene (ABS) tubes, which were made by means of a 3D printer. During the experiments, the SFCL was immersed in an open bath cryostat with liquid nitrogen and several measurements were carried out. The average limited current is about three times lower than the prospective one. To test the SFCL recovery time, the following procedure was adopted: The fault is induced during three cycles, and after, the current is reduced to zero. The SFCL is reconnected after some time considering that the fault was extinct, and the voltage level returned to normal operation. The results showed promising levels of limitation and recovery time for further developments.

Several dozen stones of an ordinary chondrite meteorite fell in and around the town of Benenitra in southwestern Madagascar during the early evening of 27 July 2018, minutes after a widely observed meteor fireball (bolide) transit and detonation. The event was confirmed by low-frequency infrasound recordings received at ~17h15 UTC (Coordinated Universal Time; 19h15 local time) at the Comprehensive Nuclear Test Ban Treaty Organization (CTBTO) infrasound station I33MG near Antananarivo, 542 km north-northeast of Benenitra. An energy release equivalent to 2.038 kt of TNT was calculated from the infrasound signals. Seismograph readings at the SKRH station 77 km north-northwest of Benenitra recorded a two-stage signal consistent with the arrivals of an initial air-coupled ground wave at 16h48:08 UTC and a stronger pulse at 16h49:22 UTC linked directly to the atmospheric pressure wave. The infrasound and seismic signal arrival times suggest that the bolide entry and detonation occurred at approximately 18h46 local time (16h46 UTC), entry was from the northwest, and the detonation hypocentre was located within ~20 km of Benenitra. Despite meteorite debris being found among buildings within Benenitra, there was no damage to structures or injuries reported. Eyewitness accounts and photographic records indicate that approximately 75 mostly intact stones were collected; however, the remoteness of the area, the rugged nature of the terrain and sales of fragments to meteorite collectors have limited scientific analysis of the fall and the extent of the strewn field. The total mass of recovered stones is estimated at between 20 kg and 30 kg, with one fragment of 11.2 kg and several of ~1 kg. Petrographic and mineral chemical analyses indicate that the stones belong to the L6 class of ordinary chondrites. Cosmogenic radionuclide analysis confirms that the fall is linked to the bolide event. The name Benenitra has been officially accepted by the Meteoritical Bulletin Database.SIGNIFICANCE: • Eyewitness reports, CTBTO infrasound records, seismograph records and cosmogenic radionuclide analysis confirm a meteorite fall over Benenitra on 27 July 2018. • Petrographic and geochemical analyses confirm that the meteorite is an L6 ordinary chondrite. • Recovery of meteorite falls is rare; this is Madagascar's second known meteorite fall and the first that can be linked to a bolide. • Regional and global science monitoring networks can be interrogated to improve the understanding of bolide events. • Interaction of scientists with local communities is important to dispel misunderstandings around scientific phenomena, and can improve collection of data.

Abstract: Skin cancer is one of the most common types of cancer in humans, it covers about one third of all neoplasms. Within skin cancer we find basal cell carcinoma (BCC), this being the most frequent type of cancer worldwide. Solutions with convolutional neural networks generally use the Softmax layer (classic model) to perform a BCC classification, however, in other similar fields such as image classification of microscopic bacteria they have replaced this Softmax layer with a support vector machine (SVM) achieving a better result. Given this, we propose a hybrid model of convolutional neural network and a support vector machine (CNN+SVM) to classify the BCC. Our model is composed of 4 convolution blocks with 32, 64 and 128 filters to carry out the extraction of characteristics and then pass it to the classifier, to which the L1-SVM loss function is implemented. The average results obtained for the CNN+SVM hybrid model were measured with the precision, accuracy, recall and F1-score metrics, obtaining 96.200%, 96.200%, 96.205% and 96.200% respectively compared to the classical model for the metrics of precision, accuracy, recall and F1-score where 95.661%, 95.673%, 95.661%, 95.660% respectively were obtained. The results show that the hybrid model achieves better results than the classic model to classify the BCC.