Letteratura scientifica selezionata sul tema "IA hybride"

Abstract This study was undertaken to determine the nature of the antigens recognized in allogeneic and syngeneic mixed leukocyte reactions (MLR). Specifically, we wished to determine whether Ia antigens alone were recognized by MLR-reactive T cells, or whether the specificity was determined by the corecognition of non-MHC antigens together with syngeneic or allogeneic Ia. To do this we used 11 T cell hybrids that were characterized as being specific for Iad and were tested their capacity to respond to isolated I-Ad or I-Ed that had been incorporated into liposomes and had bound to the surface of glass beads. Of nine alloreactive T cell hybrids (five I-Ad-and four I-Ed-specific), seven were shown to be responsive to the relevant isolated Ia antigen on glass beads. Also, two of two syngeneic I-Ad-specific T cell hybrids responded to I-Ad on the glass beads. One of the two alloreactive T cell hybrids that failed to respond to the relevant Ia antigen on glass beads was shown to be specific for an antigen in fetal calf serum (FCS) that was recognized in the context of the allo-Ia antigen (I-Ed), because when intact accessory cells were used, a response by this hybrid was only observed when FCS was present in the assay culture medium or when the accessory cells were pre-pulsed with FCS. The possible involvement of FCS antigens and non-Ia accessory cell antigens in the stimulation of the nine T cell hybrids that responded to isolated Ia on glass beads was evaluated. T cell hybrids that were grown and were tested in serum free medium were still capable of reacting to Ia on beads. The isolated Ia preparations used were greater than 90% pure, and their capacity to stimulate the T cell hybrids did not correlate with the degree of contamination with non-Ia proteins. We conclude from these studies that the majority of T cells that respond to allogeneic or syngeneic Ia bearing stimulator cells are specific for the Ia antigens themselves, and do not require the co-recognition of other non-Ia antigens; nor is there any requirement for Ia antigen processing for this recognition.

Abstract In a previous series of studies, we had shown that the constitutive Ia expression in an immunoselected Ia-human B cell variant, RJ 2.2.5, could be restored by somatic cell hybridization with mouse B cells. These experiments allowed us to show the existence of a transacting activator factor(s) operating across species barriers and encoded by the aIr-1 locus located on mouse chromosome 16. The aim of the present study was to investigate whether the B cell constitutive Ia expression and the inducible Ia expression, as seen in macrophages treated with IFN-gamma, are controlled by similar intracellular factors. To this purpose, we constructed an interspecies somatic cell hybrid between the human Ia-RJ 2.2.5 B cells and the mouse Ia-P388 D1 macrophage cells. These murine cells transiently express Ia antigens when incubated with IFN-gamma. Our results show that RJ 2.2.5 X P388 D1 cell hybrids do not express either human or mouse class II gene products. Treatment with human recombinant IFN-gamma did not modify the MHC phenotype of either the hybrid cells or the human parental cells. On the other hand, treatment of the hybrid cells with murine recombinant IFN-gamma resulted in de novo expression of mouse Ia mRNA and corresponding cell surface antigens without, however, reinduction of the human class II-positive phenotype. Furthermore, treatment with the mouse lymphokine significantly increased the levels of human HLA class I mRNA and corresponding cell surface antigens in the hybrid cells, further reinforcing the notion of the existence of non-species-specific secondary mediators generated after receptor-ligand interaction in the IFN-gamma system. Together, these results indicate that in macrophages, the intracellular events taking place after binding of IFN-gamma with its own receptor and leading to the expression of a class II-positive phenotype do not operate via an activation of the aIr-1 locus and/or its products. Thus, at least in our experimental system, we can firmly establish a first, relevant distinction between constitutive and inducible class II gene expression. This difference, dictated by the specific differentiation program of each cell type, may be relevant for the understanding of the function of class II gene products.

We report the design and synthesis of novel hydroxamic acid-tethered organoselenium (OSe) hybrids. Their antimicrobial and anticancer activities were assessed against different microbes (e.g., Candida albicans (C. albicans), Escherichia coli (E. coli), and Staphylococcus aureus (S. aureus)), as well as liver and breast carcinomas. OSe hybrid 8 showed promising anticancer activity, with IC50 = 7.57 ± 0.5 µM against HepG2 and IC50 = 9.86 ± 0.7 µM against MCF-7 cells. Additionally, OSe compounds 8 and 15 exhibited promising antimicrobial activities, particularly against C. albicans (IA% = 91.7 and 83.3) and S. aureus (IA% = 90.5 and 71.4). The minimum inhibitory concentration (MIC) assay confirmed the potential antimicrobial activity of OSe compound 8. OSe compounds 8 and 16 displayed good antioxidant activities compared to vitamin C in the 2,2-diphenyl-1-picrylhydrazyl (DPPH) and the 2,2′-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid (ABTS) assays. These results indicate that hydroxamic acid-based organoselenium hybrids have promising biological activities such as anticancer, antimicrobial, and antioxidant properties, especially compounds 8, 13, 15, and 16, which warrant further studies.

Abstract In addition to their role as peptide binding proteins, MHC class II proteins can also function as signal transducing molecules. Recent work using B cells expressing genetically engineered truncated MHC class II molecules has suggested that signaling through the cytoplasmic domains of these proteins plays an important role in the generation of signals required for the activation of some T cell hybrids. Treatment of truncated Ia-expressing B cells with cAMP-elevating agents corrects the deficiency in Ag presentation by these cells. We report that the MHC class II-mediated signal appears to act by a mechanism that increases the efficiency of Ag presentation by B cells thereby lowering the amount of specific Ag required for T cell activation. We further show that the induction of the cAMP-induced signal in B cells is inhibited by cycloheximide and cytochalasin A, implicating protein synthesis as well as cytoskeletal rearrangements in Ag presentation to accessory signal- dependent hybrids. In contrast, these agents do not block Ag presentation to a T cell hybrid previously shown not to require the cAMP-induced signal for activation. The signal-dependent T hybrid is additionally dependent on LFA-1-ICAM-1 interaction for activation, whereas the signal-independent hybrid is not. These observations suggest the existence of two types of T cell hybrid with respect to their requirements for activation: those that require only the recognition of MHC class II-peptide complexes without accessory signals, as shown by their ability to respond to purified Ia on planar membranes, and those that, in addition to recognition of MHC II/Ag, require LFA-1-ICAM-1 interaction and the delivery of additional signal(s) induced in the B cell via signal transduction through MHC class II molecules.

Inference as the basic mechanism of thought is one of the gifted abilities of human beings. It is recognized that a coherent theory and mathematical means are needed for dealing with formal causal inferences. This paper presents a novel denotational mathematical means for formal inferences known as Inference Algebra (IA). IA is structured as a set of algebraic operators on a set of formal causations. The taxonomy and framework of formal causal inferences of IA are explored in three categories: a) Logical inferences on Boolean, fuzzy, and general logic causations; b) Analytic inferences on general functional, correlative, linear regression, and nonlinear regression causations; and c) Hybrid inferences on qualification and quantification causations. IA introduces a calculus of discrete causal differential and formal models of causations; based on them nine algebraic inference operators of IA are created for manipulating the formal causations. IA is one of the basic studies towards the next generation of intelligent computers known as cognitive computers. A wide range of applications of IA are identified and demonstrated in cognitive informatics and computational intelligence towards novel theories and technologies for machine-enabled inferences and reasoning.

Community detection is one of the most challenging and interesting problems in many research areas. Being able to detect highly linked communities in a network can lead to many benefits, such as understanding relationships between entities or interactions between biological genes, for instance. Two different immunological algorithms have been designed for this problem, called Opt-IA and Hybrid-IA, respectively. The main difference between the two algorithms is the search strategy and related immunological operators developed: the first carries out a random search together with purely stochastic operators; the last one is instead based on a deterministic Local Search that tries to refine and improve the current solutions discovered. The robustness of Opt-IA and Hybrid-IA has been assessed on several real social networks. These same networks have also been considered for comparing both algorithms with other seven different metaheuristics and the well-known greedy optimization Louvain algorithm. The experimental analysis conducted proves that Opt-IA and Hybrid-IA are reliable optimization methods for community detection, outperforming all compared algorithms.

Inference as the basic mechanism of thought is abilities gifted to human beings, which is a cognitive process that creates rational causations between a pair of cause and effect based on empirical arguments, formal reasoning, and/or statistical norms. It’s recognized that a coherent theory and mathematical means are needed for dealing with formal causal inferences. Presented is a novel denotational mathematical means for formal inferences known as Inference Algebra (IA) and structured as a set of algebraic operators on a set of formal causations. The taxonomy and framework of formal causal inferences of IA are explored in three categories: a) Logical inferences; b) Analytic inferences; and c) Hybrid inferences. IA introduces the calculus of discrete causal differential and formal models of causations. IA enables artificial intelligence and computational intelligent systems to mimic human inference abilities by cognitive computing. A wide range of applications of IA are identified and demonstrated in cognitive informatics and computational intelligence towards novel theories and technologies for machine-enabled inferences and reasoning. This work is presented in two parts. The inference operators of IA as well as their extensions and applications will be presented in this paper; while the structure of formal inference, the framework of IA, and the mathematical models of formal causations has been published in the first part of the paper in IJCINI 5(4).

Abstract In this article, we analyse the factors underpinning the shift towards hybrid security governance in Africa. Extant scholarship largely attributes this shift to broader global processes, such as histories of colonialism, neoliberalism and transformations in global governance, which have served to legitimize the role of private authority in security provision around sites of resource extraction. Our analysis seeks to understand the relative and relational influence of power and rules in international politics by offering empirical insights about what hybrid security arrangements look like ‘on the ground’. Drawing upon recently conducted fieldwork in Kenya, Uganda and Ghana, we examine how hybrid security arrangements affect the lives of those living near sites of natural resource extraction. Our analyses suggest that although hybrid security has emerged as the leading approach to security governance, this approach to security does not uniformly involve or serve the interests of all stakeholders. Rather, we find that hybrid security arrangements aid the security of extractive operations—securing investments in both physical and human capital—while sometimes undermining the security of nearby communities.

La vision par ordinateur a connu une évolution importante, passant des méthodes traditionnelles aux modèles d'apprentissage profond. L’un des principaux objectifs des tâches de vision par ordinateur est d’émuler la perception humaine. En effet, le processus classique effectué par les modèles d’apprentissage profond dépend entièrement des caractéristiques visuelles, reflétant simplement la manière dont les humains perçoivent visuellement leur environnement. Cependant, pour que les humains comprennent l’environnement qui les entoure, leur raisonnement dépend non seulement de leurs capacités visuelles, mais aussi de leurs connaissances pré-acquises. Combler cette différence entre la perception humaine et celle des machines est essentielle afin de parvenir à un raisonnement similaire à celui des humains. Dans cette thèse, nous proposons de nouvelles approches pour améliorer les performances des modèles d’apprentissage profond en intégrant les systèmes basés sur les connaissances dans les réseaux de neuronaux profonds. L'objectif est d’aider ces réseaux à prendre les bonnes décisions en exploitant à la fois les caractéristiques visuelles et les connaissances pour émuler l’analyse visuelle de l’être humain. Ces méthodologies impliquent deux axes principaux. Premièrement, définir la représentation des connaissances pour incorporer des informations utiles à une tâche spécifique de vision. Deuxièmement, examiner comment intégrer ces connaissances dans les réseaux de neurones pour améliorer leurs performances. La première contribution porte sur l'estimation de la profondeur monoculaire. En effet, les humains sont capables d'estimer leur distance par rapport aux objets perçus, même en n’utilisant qu’un seul œil, et ceci en se basant sur les indices monoculaires. Nous proposons d'intégrer ces indices au sein des réseaux de neurones comme un raisonnement similaire à celui des humains pour l'estimation de la profondeur. À cette fin, nous suggérons d'exploiter un modèle ontologique pour représenter l'environnement comme un ensemble de concepts liés par des relations sémantiques. Les informations sur les indices monoculaires sont extraites grâce à un raisonnement effectué sur l'ontologie proposée et sont transférées dans les réseaux de neurones. Le deuxième travail porte sur la tâche de segmentation panoptique qui vise à identifier toutes les instances d’objets capturées dans une image. Nous proposons une approche qui combine les avantages des réseaux de neurones avec des connaissances sur les relations spatiales entre les objets. Nous avons choisi ce type de connaissances car elles peuvent fournir des indices utiles pour résoudre les ambiguïtés et distinguer entre les instances d'objets similaires. Plus précisément, nous proposons une stratégie d'entraînement qui intègre les connaissances dans le processus d'optimisation des réseaux de neurones. L’approche comprend un processus d'extraction et de représentation des connaissances sur les relations spatiales, qui sont incorporées dans l’entraînement sous forme d'une fonction de perte. Afin de valider l'efficacité des approches proposées, nous avons choisi l'environnement urbain et les véhicules autonomes comme principale cas d’application. Ce domaine est particulièrement intéressant car il s'agit d'un axe de recherche novateur en développement continu, avec des implications significatives pour la sécurité et la mobilité des humains. En conclusion, nous avons étudié diverses approches pour représenter les connaissances et les intégrer aux réseaux de neurones. Ces approches valident que l’utilisation combinée de méthodes basées sur les connaissances et celles basées sur les données conduit de manière constante à des résultats améliorés. Le défi principal réside toujours dans le choix des connaissances pertinentes pour chaque tâche, leur représentation et leur intégration de la manière la plus optimale dans l'architecture des réseaus de neurones profonds
Computer vision has made an important evolution starting from traditional methods to advanced Deep Learning (DL) models. One of the goals of computer vision tasks is to effectively emulate human perception. The classical process of DL models is completely dependent on visual features, which only reflects how humans visually perceive their surroundings. However, for humans to comprehensively understand their environment, their reasoning not only depends on what they see but also on their pre-acquired knowledge. Addressing this gap is essential as achieving human-like reasoning requires a seamless combination of data-driven and knowledge-driven methods. In this thesis, we propose new approaches to improve the performance of DL models by integrating Knowledge-Based Systems (KBS) within Deep Neural Networks (DNNs). The goal is to empower these networks to make informed decisions by leveraging both visual features and knowledge to emulate human-like visual analysis. These methodologies involve two main axes. First, define the representation of KBS to incorporate useful information for a specific computer vision task. Second, investigate how to integrate this knowledge into DNNs to enhance their performance. To do so, we worked on two main contributions. The first work focuses on monocular depth estimation. Considering humans as an example, they can estimate their distance with respect to seen objects, even using just one eye, based on what is called monocular cues. Our contribution involves integrating these monocular cues as human-like reasoning for monocular depth estimation within DNNs. For this purpose, we investigate the possibility of directly integrating geometric and semantic information into the monocular depth estimation process. We suggest using an ontology model in a DL context to represent the environment as a structured set of concepts linked with semantic relationships. Monocular cues information is extracted through reasoning performed on the proposed ontology and is fed together with the RGB image in a multi-stream way into the DNNs. Our approach is validated and evaluated on widespread benchmark datasets. The second work focuses on panoptic segmentation task that aims to identify and analyze all objects captured in an image. More precisely, we propose a new informed deep learning approach that combines the strengths of DNNs with some additional knowledge about spatial relationships between objects. We have chosen spatial relationships knowledge for this task because it can provide useful cues for resolving ambiguities, distinguishing between overlapping or similar object instances, and capturing the holistic structure of the scene. More precisely, we propose a novel training methodology that integrates knowledge directly into the DNNs optimization process. Our approach includes a process for extracting and representing spatial relationships knowledge, which is incorporated into the training using a specially designed loss function. The performance of the proposed method was also evaluated on various challenging datasets. To validate the effectiveness of the proposed approaches for combining KBS and DNNs regarding different methodologies, we have chosen the urban environment and autonomous vehicles as our main use case application. This domain is particularly interesting because it is a challenging and novel field in continuous development, with significant implications for the safety, comfort and mobility of humans. As a conclusion, the proposed approaches validate that the integration of knowledge-driven and data-driven methods consistently leads to improved results. Integration improves the learning process for DNNs and enhances results of computer vision tasks, providing more accurate predictions. The challenge always lies in choosing the relevant knowledge for each task, representing it in the best structure to leverage meaningful information, and integrating it most optimally into the DNN architecture

Talent is increasingly important for organizations which makes corporate recruitment a significant ongoing function. Recruiting top talent cannot occur where there are inefficiencies, high costs and lack of innovation. At the same time, artificial intelligence (AI) and machine learning (ML) are disrupting industries and different areas of business practice. This technology has the potential to create unprecedented value in recruiting functions also, by positively impacting efficiency, costs and employee fit. Despite rapid developments in the field of AI, academic literature on AI in recruitment is thin. Researchers further call for more collaborative work between practitioners and academics. This thesis aims to tackle this gap, evaluating how AI and ML disrupt traditional recruitment processes and bring about potential new outcomes. By integrating the experiences of experts, executives and the perceptions of potential job applicants, this thesis elucidates practical implications for the adoption of AI and ML in recruiting. The thesis uses qualitative and quantitative data collection. The findings present opportunities and also the limitations of AI and ML in recruiting. Furthermore, the effects of the technology on recruiting efficiency and validity are evaluated. This creates the basis for practical recommendations for organizations regarding the adoption of this technology. Notably, in the more standardized aspects of the recruitment processes, this technology does create value in hiring.
O talento é cada vez mais importante para as organizações que utilizam o recrutamento corporativo como uma função contínua e significativa. O recrutamento dos melhores talentos não pode ocorrer onde existem ineficiências, altos custos e falta de inovação. Ao mesmo tempo, a inteligência artificial (IA) e machine learning (ML) estão rompendo indústrias e diferentes áreas de prática de negócios. Essa tecnologia tem o potencial de criar um valor sem precedentes nas funções de recrutamento, impactando positivamente a eficiência, os custos e a adequação dos funcionários. Apesar do rápido desenvolvimento no campo da IA, a literatura acadêmica sobre IA no recrutamento é escassa. Os pesquisadores gostariam que existisse mais trabalho colaborativo entre profissionais e acadêmicos. Esta tese visa abordar essa lacuna, avaliando como a IA e o ML modificam os processos tradicionais de recrutamento e trazem novos resultados potenciais. Ao integrar as experiências de especialistas, executivos e as percepções de possíveis candidatos a emprego, esta tese elucida implicações práticas para a adoção de IA e ML no recrutamento. A tese utiliza coleta de dados qualitativa e quantitativa. Os resultados apresentam oportunidades e também as limitações da IA e ML. Além disso, os efeitos da tecnologia no recrutamento eficiente e válido são avaliados. Isso cria a base para recomendações práticas para as organizações com relação à adoção desta tecnologia. Notavelmente, nos aspectos mais padronizados dos processos de recrutamento, essa tecnologia cria valor na contratação.

In this paper, an action mechanism of the carbon emission intensity (CE) reduction relevant to the level of industrial agglomeration (IA), scientific and technological progress (TECH), foreign direct investment (F), and environmental regulation (ER) is studied by applying a hybrid framework driven by the K-means clustering algorithm and the stepwise multiple linear regression (SMLR)-based models. To concisely clarify the relationship between the objected variable and impact factors, three echelons of urban agglomerations are summarized in this study from the different industry characteristics of the whole 21 cities in Guangdong, which exhibits a strong growth of the secondary industry in China. Stepwise multiple linear regression (SMLR) analysis is proposed to reveal that strict environmental regulations for the urban agglomeration with low overall development level and weak industrial foundation are more likely to stimulate the “innovation compensation” effect. For the urban agglomeration in the growth stage of industrial economic development or distributed in the developed region, strengthening the level of IA and highlighting the core guiding role of environmental regulation seems to make more sense to reduce the carbon emission intensity and promote the regional industrial green upgrading.

Optical memory systems utilizing volume holographic storage (VHS) have received considerable attention in recent years[1-3]. One aspect of VHS is the ability to copy stored data from an archive (a VHS medium containing N permanently-stored data sets) to a secondary storage medium (SSM). To achieve the copying of volume holographic memories from the archive to the SSM, three primary approaches can be utilized: parallel[4], incremental[5], and sequential[6]. In the parallel approach, there exist N mutually incoherent read beams that enter the archive and read out and copy all N stored data sets simultaneously. In the incremental approach, there exists a single coherent read beam that is rapidly multiplexed over time (in angle, wavelength, phase, et cetera) in a manner such that it reads out and copies all N data sets serially, over many repeated iterations in which each iteration time is short compared to the SSM’s response time. In the sequential approach, there also exists a single coherent read beam that is multiplexed over time (in angle, wavelength, phase, et cetera) but such that it reads out and copies the N data sets sequentially in a manner that follows an appropriate (single-pass) exposure schedule. In this paper, we explore the fundamental material limitations in each of these copying approaches for the cases of all-optical, quasi all-optical, and hybrid opto-electronic copying schemes. These limitations include the maximum allowable intensity, IA, into the archive that will not damage it, the resulting total intensity, ID, of the diffracted beam from the archive as a function of N, IA, and other archival material and geometry constraints, dark conductivity, shot and Kerr noise in the SSM, cross-talk, and the characteristics of any devices that may be placed between the archive and the SSM.

Abstract An investigation into the computation of hydrodynamic loads on a suspended cylinder in regular waves is presented. The primary goal was to perform a three-way validation of the loads between experimental measurements and simulations from two computational methods. Experimental measurements of the longitudinal in-line force on a cylinder suspended at a fixed position were available from the Offshore Code Comparison Collaboration, Continued, with Correlation (OC5) project, Phase Ia. These measurements were compared to computational fluid dynamics (CFD) simulations based on the solution of Reynolds-averaged Navier-Stokes (RANS) equations, as implemented in STAR-CCM+. The study encompassed a sensitivity analysis of the loads computed in STAR-CCM+ based on wave modeling, boundary conditions, turbulence modeling, and spatial and temporal discretization. The analysis was supplemented by results generated with the offshore wind turbine engineering software OpenFAST, based on a hybrid combination of second-order potential flow and viscous drag from Morison’s equation. The focus of the investigation was on the assessment of the accuracy of the computation of first- and higher-order hydrodynamic loads. Substantial differences were observed in the numerical prediction of the second and third harmonic force contribution. Local flow field analysis with CFD was applied to study the physics of wave run-up and diffraction dynamics to identify the causes.