Search

Submitted by Cesar Andrade (goersch@hotmail.com) on 2021-04-16T22:59:28Z No. of bitstreams: 4 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Dissertacao_Cesar_Henrique_Goersch_Andrade.pdf: 1183949 bytes, checksum: a0358cd3aa82363019e8d0c73841b457 (MD5) folhaaprov.pdf: 969267 bytes, checksum: 24d280bbaf9f356691383122cab79be9 (MD5) Carta de Autoriza????o de Encaminhamento Cesar Andrade.pdf: 147190 bytes, checksum: 3ef4c8925c0fc3b576710b2f73111346 (MD5) Approved for entry into archive by PPGI Inform??tica (secretariappgi@icomp.ufam.edu.br) on 2021-04-27T16:58:10Z (GMT) No. of bitstreams: 4 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Dissertacao_Cesar_Henrique_Goersch_Andrade.pdf: 1183949 bytes, checksum: a0358cd3aa82363019e8d0c73841b457 (MD5) folhaaprov.pdf: 969267 bytes, checksum: 24d280bbaf9f356691383122cab79be9 (MD5) Carta de Autoriza????o de Encaminhamento Cesar Andrade.pdf: 147190 bytes, checksum: 3ef4c8925c0fc3b576710b2f73111346 (MD5) Approved for entry into archive by Divis??o de Documenta????o/BC Biblioteca Central (ddbc@ufam.edu.br) on 2021-04-28T00:06:13Z (GMT) No. of bitstreams: 4 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Dissertacao_Cesar_Henrique_Goersch_Andrade.pdf: 1183949 bytes, checksum: a0358cd3aa82363019e8d0c73841b457 (MD5) folhaaprov.pdf: 969267 bytes, checksum: 24d280bbaf9f356691383122cab79be9 (MD5) Carta de Autoriza????o de Encaminhamento Cesar Andrade.pdf: 147190 bytes, checksum: 3ef4c8925c0fc3b576710b2f73111346 (MD5) Made available in DSpace on 2021-04-28T00:06:13Z (GMT). No. of bitstreams: 4 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Dissertacao_Cesar_Henrique_Goersch_Andrade.pdf: 1183949 bytes, checksum: a0358cd3aa82363019e8d0c73841b457 (MD5) folhaaprov.pdf: 969267 bytes, checksum: 24d280bbaf9f356691383122cab79be9 (MD5) Carta de Autoriza????o de Encaminhamento Cesar Andrade.pdf: 147190 bytes, checksum: 3ef4c8925c0fc3b576710b2f73111346 (MD5) Previous issue date: 2021-03-18 The personal and corporative computers predominantly use accounts credentials (for example: login and password) as an authentication method, also known as statistic methods. A problem with this approach is that the user can leave the computer without logout or lock the access and allowing an intruder to access the available resources. Therefore, recently searches are directing its efforts into continuing authentication solutions, based on the user???s behavioral models. Most of the approaches uses authentication models built from information that are extracted from user???s interactions with the devices, such as, mouse movements or dynamic text typing or speech recognition. Different form the existing approaches, this work has the propose to make use of static information related to the use of the hardware and software, which are obtained from the performance???s counter from the operational systems to generate authentication models. The idea is to take the information related to the usage of computers sources by the user over time as the use of memory, processor, network, storage and application, to create a profile that can be used to authenticate the user. The advantage to use those attributes is that they can be collected in a transparent way without interfering on the user???s activity. Be sides, the main operational systems (for example, Linux and Windows) have already made available native collectors, not requiring the development of specific collection software. To generate de automation models, we used a hybrid deep network architecture, composed by convolution layers and by recurrence layer. The convolution layers perform the automatic extraction of the characteristics (in this case, the correlation between the data from the performance counter) and the recurrence layer are used to computed temporal characteristic from the data processed by the convolution layers. Furthermore, this work employs a trust model that avoids blocking genuine users and prevents an imposter from spending too much time undetected. The results obtained in three evaluation scenarios show that the proposed method can detect 100% of imposter users in up to 15 seconds. These results prove the feasibility of using performance counters in the definition of continuous authentication models. Os computadores pessoais e coorporativos predominantemente utilizam credenciais de contas (e.g. login e senha) como m??todo de autentica????o, tamb??m conhecidos como m??todos est??ticos. Um problema com esta abordagem ?? que o usu??rio pode deixar o computador sem sair da sess??o ou bloquear seu acesso, possibilitando a um intruso acessar os recursos dispon??veis. Por essa raz??o, pesquisas recentes t??m direcionado seus esfor??os em solu????es de autentica????o cont??nua baseada em modelos comportamentais dos usu??rios. A maioria das abordagens emprega modelos de autentica????o constru??dos a partir de informa????es extra??das das intera????es dos usu??rios com os dispositivos como, por exemplo, a partir de movimentos do mouse, din??mica na digita????o de textos ou reconhecimento de fala. Diferentemente das abordagens existentes, este trabalho prop??e a utiliza????o de informa????es estat??sticas relacionadas ao uso de hardware e software obtidos a partir dos contadores de desempenho dos sistemas operacionais para gerar modelos de autentica????o. A ideia ?? usar as informa????es relacionadas ao uso dos recursos de um computador pelo usu??rio ao longo do tempo como o uso de mem??ria, processador, rede, armazenamento e aplica????es, para criar um perfil que possa ser usado para autenticar o usu??rio. A vantagem do uso destes atributos ?? que eles podem ser coletados de forma transparente, sem interferir na atividade do usu??rio. Al??m disso, os principais sistemas operacionais (e.g. Linux e Windows) j?? disponibilizam coletores nativos, n??o requerendo o desenvolvimento de softwares de coleta espec??ficos. Para gerar os modelos de autentica????o, n??s empregamos uma arquitetura de rede profunda h??brida, composta por camadas de convolu????o e por camadas de recorr??ncia. As camadas de convolu????o realizam a extra????o autom??tica de caracter??sticas (neste caso, correla????es entre dados dos contadores de desempenho) e as camadas de recorr??ncia s??o utilizadas para capturar caracter??sticas temporais dos dados processado pelas camadas convolucionais. Al??m disso, este trabalho emprega um modelo de confian??a que evita o bloqueio de usu??rios genu??nos e impede que um impostor fique muito tempo agindo sem ser detectado. Os resultados obtidos em tr??s cen??rios de avalia????o mostram que o m??todo proposto consegue detectar 100% dos usu??rios impostores em at?? 15 segundos. Os resultados comprovam a viabilidade do uso de contadores de desempenho na defini????o de modelos de autentica????o cont??nua.

Chemical production has continued to in-crease and change since 2003. The potential geographic impact of the chemical industry continued to change from the Atlantic Ocean to the Pacific Ocean, where almost 70 per cent of the industry is expected to operate by 2030, while new products are continually being devel-oped, thus adding to the mixture of chemicals to which biota in the ocean is being exposed.Different lists of hazardous substances have been identified by international organizations, although there is still no agreed single global list of substances that are of concern. The present chapter contains an assessment of the changes since the first Assessment in wa-ter and airborne inputs to the marine environ-ment from land (including groundwater), ships and offshore installations. In addition, the information in the present chapter builds upon the assessment of the list of hazardous sub-stances used in the first Assessment, namely, POPs, metals, hydrocarbons and radioactive substances. It includes new information on rare earth elements, PPCPs and airborne in-puts of nitrogen oxides and sulfur oxides that were not included in the first Assessment.

With the increasing adoption of teledermatology, there is a need to improve the automatic organization of medical records, being dermatological image modality a key filter in this process. Although there has been considerable effort in the classification of medical imaging modalities, this has not been in the field of dermatology. Moreover, as various devices are used in teledermatological consultations, image acquisition conditions may differ. In this work, two models (VGG-16 and MobileNetV2) were used to classify dermatological images from the Portuguese National Health System according to their modality. Afterwards, four incremental learning strategies were applied to these models, namely naive, elastic weight consolidation, averaged gradient episodic memory, and experience replay, enabling their adaptation to new conditions while preserving previously acquired knowledge. The evaluation considered catastrophic forgetting, accuracy, and computational cost. The MobileNetV2 trained with the experience replay strategy, with 500 images in memory, achieved a global accuracy of 86.04% with only 0.0344 of forgetting, which is 6.98% less than the second-best strategy. Regarding efficiency, this strategy took 56 s per epoch longer than the baseline and required, on average, 4554 megabytes of RAM during training. Promising results were achieved, proving the effectiveness of the proposed approach.

Background ESGO (European Society of Gynaecological Oncology) and partners are continually improving the developmental opportunities for gynaecological oncology fellows. The objectives of this survey were to evaluate the progress in the infrastructure of the training systems in Europe over the past decade. We also evaluated training and assessment techniques, the perceived relevance of ENYGO (European Network of Young Gynaecological Oncologists) initiatives, and unmet needs of trainees. Methodology National representatives of ENYGO from 39 countries were contacted with an electronic survey. A graduation in well/moderately/loosely-structured training systems was performed. Descriptive statistical analysis and frequency tables, as well as two-sided Fisher's exact test, were used. Results National representatives from 33 countries answered our survey questionnaire, yielding a response rate of 85%. A national fellowship is offered in 22 countries (66.7%). A logbook to document progress during training is mandatory in 24 (72.7%) countries. A logbook of experience is only utilized in a minority of nations (18%) for assessment purposes. In 42.4% of countries, objective assessments are recognized. Trainees in most countries (22 (66.7%)) requested additional training in advanced laparoscopic surgery. 13 (39.4%) countries have a loosely-structured training system, 11 (33.3%) a moderately-structured training system, and 9 (27.3%) a well-structured training system. Conclusion Since the last publication in 2011, ENYGO was able to implement new activities, workshops, and online education to support training of gynaecological oncology fellows, which were all rated by the respondents as highly useful. This survey also reveals the limitations in establishing more accredited centers, centralized cancer care, and the lack of laparoscopic training. © 2021 BMJ Publishing Group. All rights reserved.

Jellyfish are an important component of pelagic ecosystems that can form massive blooms over large spatial and temporal scales. Intensity, frequency, and duration of jellyfish blooms vary, show strong species- and site-specific differences and are often considered as nuisance for local industries, tourism and artisanal fishery. Demands for jellyfish-related products in e.g. the aquaculture feed, foods, cosmetics or pharmaceutical sector are increasing becoming concrete specially for food uses of some species, thus providing an opportunity to turn this nuisance into a resource. However, a sound management of jellyfish resources is missing thus hampering a sustainable, cost-efficient harvesting of jellyfish biomass. Within the framework of the EU-project GoJelly, the aim was to provide a basis for ecosystem-based management approaches taking the knowledge gaps related to jellyfish ecology, bloom dynamics, and origin of seed populations into account. By focusing on the biotic and abiotic drivers that promote jellyfish bloom formation, more reliable predictions on jellyfish blooms and tools for sustainable harvesting should be achieved. From 2018-2020, target jellyfish species with broad spatiotemporal distribution ranges in European waters were harvested (Aurelia spp., Periphylla periphylla, Cyanea capillata, Rhopilema nomadica, Rhizostoma pulmo, Cotylorhiza tuberculata, Catostylus tagi, Mnemiopsis leidyi) in several GoJelly case areas (Norwegian Sea, Baltic Sea, Mediterranean Sea and Eastern Atlantic). Different types of fishing vessels were used ranging from small-sized open boats to medium-sized research vessels. Jellyfish were harvested using six gear types (gillnets, bottom trawls, pelagic trawls, landing nets, hand nets, 20-L buckets). The mesh size varied from 0.35 to 50 mm. The gear type reflected the harvesting depth as well as the vessel type. In total approx. 3 tons of jellyfish were harvested by GoJelly partners whereof 60% of the total biomass was harvested in the Trondheimsfjord case area, Norway. The harvested biomass showed interannual variation related to differences in harvesting efforts and bloom intensities of the target species (Figure 2A). Nine jellyfish species were harvested (Figure 2B), three on them representing 80% of the total catch (44% Periphylla periphylla: harvested in Trondheimsfjord; 18% Rhizostoma pulmo: harvested in the Mediterranean Sea (Gulf of Trieste and Ionian Sea); 17% Aurelia spp.: harvested in Trondheimsfjord, Baltic Sea and Gulf of Trieste (Mediterranean). Management of jellyfish resources in Europe is currently missing thus hampering a sustainable, cost-efficient harvesting of jellyfish biomass. Within the EU-project GoJelly, the aim was to provide a basis for ecosystem-based management approaches taking the knowledge gaps related to jellyfish ecology, bloom dynamics, and origin of seed populations into account. The aim was to analyse biotic and abiotic drivers that promote jellyfish bloom formation in the Trondheimsfjord case area to enable more reliable predictions on jellyfish blooms and tools for sustainable harvesting. Here, we address the knowledge gaps related to jellyfisheries and demonstrate future needs towards an ecosystem-based management of local jellyfish resources.

Benthic foraminifera from twenty stations sampled twice were analyzed in order to examine the environmental conditions of the Albufeira coastal lagoon. Foraminiferal assemblages show an increase in the abundance, species diversity and allochthonous content seaward. Three zones have been defined: the inner and more restricted area of the lagoon where the main species are the euryhaline Ammonia tepida and Haynesina germanica; these two species together with Bulimina gibba and Brizalina britannica dominated the middle area; and, finally, in the mouth and marine- influenced back-barrier area the assemblages are made up of brackish and marine species such as A. tepida, Cribroelphidium williamsoni, Cibicides lobatulus and H. germanica. Additionally, samples from three cores were also examined. Sedimentological and micropalaeontological contents represent the last 2500 years of environmental evolution of the lagoon. Foraminiferal assemblages are not very abundant and they are dominated by H. germanica, A. tepida and Jadammina macrescens indicating very restricted conditions through time. In a regional context, and during the Late Holocene, after the development of a detrital barrier circa 5.5 cal yr BP, eustasy is considered to be a minor environmental control of the barrier-lagoonal system dynamics. Instead, environmental conditions were forced by local factors, such as frequency and duration of lagoon-ocean water exchanges, induced by both natural and artificial changes in the permeability of the sand-barrier.

Coelomocytes is a generic name for a collection of cellular morphotypes, present in many coelomate animals, that has been reported as highly variable across echinoderm classes. The roles attributed to the major types of the free circulating cells present in the coelomic fluid of echinoderms include immune response, phagocytic digestion and clotting. The main aim of the present study is the thorough characterization of coelomocytes present in the coelomic fluid ofMarthasterias glacialis(class Asteroidea) through the combined use of flow cytometry (FC) and fluorescence plus transmission electron microscopy. Two coelomocyte populations (here named P1 and P2) were identified by flow cytometry and subsequently studied in terms of abundance, morphology, ultrastructure, cell viability and cell cycle profiles. Ultrastructurally, P2 diploid cells showed two main morphotypes, similar to phagocytes and vertebrate thrombocytes, whereas the small P1 haploid cellular population was characterized by a low mitotic activity, relatively undifferentiated cytotype and a high nucleus/cytoplasm ratio. These last cells resemble stem-cell types present in other animals. P1 and P2 cells differ also in cell viability and cell cycle profiles. Additionally, two other morphotypes were only detected by fluorescence microscopy and a third one when using a combined microscopy/FC approach.

Bond between steel and concrete in prestressed concrete members may be degraded due to different reasons, being one of the most frequent ones the corrosion induced damage. Transfer length is the key indicator to describe the bond properties between prestressed steel and concrete. Through tests on prestressed beams, this paper is focused on the experimental verification of an analytical model developed to analyse the processes of bond degradation due to the corrosion of both indented and smooth wires, and its influence on the behaviour of prestressed concrete members. The model is based on both the thick-walled cylinder theory and the cohesive cracking for concrete. En elementos pretensados de hormigón, la adherencia entre el acero y el hormigón puede verse deteriorada por diferentes motivos, siendo uno de los más recurrentes el daño inducido por procesos de corrosión. El parámetro fundamental para caracterizar las propiedades adherentes existentes entre el acero pretensado y el hormigón es la longitud de transferencia. Este artículo compara los resultados experimentales obtenidos en ensayos sobre elementos pretensados tipo viga con las predicciones arrojadas por un modelo analítico, basado en la teoría de cilindros de pared gruesa y en la fisuración cohesiva del hormigón, desarrollado para analizar los procesos de degradación de la adherencia motivados por la corrosión en alambres lisos e indentados y su influencia en el comportamiento de elementos pretensados de hormigón.