Добірка наукової літератури з теми "Pseudo-High-Entropy"

The rapid penetration of the banking market of mobile banking creates not only new opportunities and conveniences, but also new risks and threats, which leads to increased requirements for information security of mobile services. One of the important components of the security of the mobile banking service of the client-bank system is the protection mechanisms based on the use of cryptographic keys. The stability of cryptographic keys is a guarantee of system reliability and, in turn, is based on the mechanisms of generating pseudo-random numbers using primary sources of entropy, for the effective use of which it is necessary to determine the type and capabilities of the source. The paper presents the results of the study of sensors of mobile communication devices as a source of entropy for generating cryptokeys of a mobile application of the client-bank system. An accelerometer, a temperature sensor, a gyroscope, a light brightness sensor, a magnetic field sensor, an atmospheric pressure sensor, a distance sensor, and a humidity sensor are considered as a source of entropy. The results of the research confirm the presence in the values of the sensors of mobile communication devices of a component suitable for use as a source of primary entropy for generators of pseudo-random numbers by cryptographic algorithms. In turn, pseudo-random number generators built with the use of mobile device sensors can be the basis of high-entropy cryptographic key generators in the client-bank system mobile banking services. The obtained results are focused on the implementation of algorithms and means of generating high-entropy pseudo-random numbers and can be used to reduce the risks of unauthorized access to customer information in the client-bank system by disclosing cryptokeys in mobile banking services.

A novel active semi-supervised learning framework using unlabeled data is proposed for fault identification in labeled expensive chemical processes. A principal component analysis (PCA) feature selection strategy is first given to calculate the weight of the variables. Secondly, the identification model is trained based on the obtained key process variables. Thirdly, the pseudo label confidence of identification model is dynamically optimized with an historical, current, and future pseudo label confidence mean. To increase the upper limit of the identification model that is self-learning with high entropy process data, active learning is used to identify process data and diagnosis fault causes by ontology. Finally, a PCA-dynamic active safe semi-supervised support vector machine (PCA-DAS4VM) for fault identification in labeled expensive chemical processes is built. The application in the Tennessee Eastman (TE) process shows that this hybrid technology is able to: (i) eliminate chemical process noise and redundant process variables simultaneously, (ii) combine historical pseudo label confidence with future pseudo label confidence to improve the identification accuracy of abnormal working conditions, (iii) efficiently select and diagnose high entropy unlabeled process data, and (iv) fully utilize unlabeled data to enhance the identification performance.

High entropy alloys (HEAs) generally exhibit either high resistance to deformation or high toughness due to the presence of body-centered or face-centered cubic structure, respectively. To overcome these limitations, new high entropy alloys have been developed in the present study. This investigation aims to synthesis and characterization of novel CoCrFeNi3Si, CoCrFe2Ni2Si, and Co2CrFeNi2Si high entropy alloys. The mechanical alloying route is used to synthesize these alloys. Grinding was carried out to 20h and X-ray diffraction (XRD) analysis was done at different time intervals of grinding. The face-centered cubic structure along with the intermetallic compound of Ni-Si was observed after 20h of grinding. Furthermore, a pseudo binary strategy based on the valence electron concentration and mixing enthalpy is also employed to design the high entropy alloys considered in the present study. Carefully analysis of the XRD pattern indicates that from 5 to 20h of mechanical alloying there is a decrement in the initial peaks of elements observed.

Hyperspectral image classification methods may not achieve good performance when a limited number of training samples are provided. However, labeling sufficient samples of hyperspectral images to achieve adequate training is quite expensive and difficult. In this paper, we propose a novel sample pseudo-labeling method based on sparse representation (SRSPL) for hyperspectral image classification, in which sparse representation is used to select the purest samples to extend the training set. The proposed method consists of the following three steps. First, intrinsic image decomposition is used to obtain the reflectance components of hyperspectral images. Second, hyperspectral pixels are sparsely represented using an overcomplete dictionary composed of all training samples. Finally, information entropy is defined for the vectorized sparse representation, and then the pixels with low information entropy are selected as pseudo-labeled samples to augment the training set. The quality of the generated pseudo-labeled samples is evaluated based on classification accuracy, i.e., overall accuracy, average accuracy, and Kappa coefficient. Experimental results on four real hyperspectral data sets demonstrate excellent classification performance using the new added pseudo-labeled samples, which indicates that the generated samples are of high confidence.

Semantic segmentation is a growing topic in high-resolution remote sensing image processing. The information in remote sensing images is complex, and the effectiveness of most remote sensing image semantic segmentation methods depends on the number of labels; however, labeling images requires significant time and labor costs. To solve these problems, we propose a semi-supervised semantic segmentation method based on dual cross-entropy consistency and a teacher–student structure. First, we add a channel attention mechanism to the encoding network of the teacher model to reduce the predictive entropy of the pseudo label. Secondly, the two student networks share a common coding network to ensure consistent input information entropy, and a sharpening function is used to reduce the information entropy of unsupervised predictions for both student networks. Finally, we complete the alternate training of the models via two entropy-consistent tasks: (1) semi-supervising student prediction results via pseudo-labels generated from the teacher model, (2) cross-supervision between student models. Experimental results on publicly available datasets indicate that the suggested model can fully understand the hidden information in unlabeled images and reduce the information entropy in prediction, as well as reduce the number of required labeled images with guaranteed accuracy. This allows the new method to outperform the related semi-supervised semantic segmentation algorithm at half the proportion of labeled images.

Unsupervised domain adaptation (UDA) is essential since manually labeling pixel-level annotations is consuming and expensive. Since the domain discrepancies have not been well solved, existing UDA approaches yield poor performance compared with supervised learning approaches. In this paper, we propose a novel sequential learning network (SLNet) for unsupervised cross-scene aerial image segmentation. The whole system is decoupled into two sequential parts—the image translation model and segmentation adaptation model. Specifically, we introduce the spectral space transferring (SST) approach to narrow the visual discrepancy. The high-frequency components between the source images and the translated images can be transferred in the Fourier spectral space for better preserving the important identity and fine-grained details. To further alleviate the distribution discrepancy, an efficient pseudo-label revising (PLR) approach was developed to guide pseudo-label learning via entropy minimization. Without additional parameters, the entropy map works as the adaptive threshold, constantly revising the pseudo labels for the target domain. Furthermore, numerous experiments for single-category and multi-category UDA segmentation demonstrate that our SLNet is the state-of-the-art.

La quête de matériaux alternatifs pour remplacer l'acier inoxydable conventionnel dans les applications marines a récemment gagné en importance. Parmi les candidats émergents, les alliages amorphes à base de fer à faible teneur en chrome ont montré un grand potentiel, bien qu'ils nécessitent une compréhension plus approfondie de leur comportement. Cette recherche étudie l'effet de la dévitrification sur la corrosion et les propriétés mécaniques de deux compositions amorphes pseudo-entropiques novatrices à base de FeCrMoNbB et FeCrMoNiB, ainsi que leur comportement lorsqu'ils sont appliqués en tant que revêtements, mettant en lumière le rôle critique de la caractérisation du film passif.Dans la première étape, des rubans préparés par refusion ont été soumis à un recuit dans une atmosphère protectrice d'argon pour simuler la cristallisation. Des tests de corrosion ont été réalisés dans une cellule à trois électrodes, révélant la sensibilité des propriétés de corrosion au processus de dévitrification, en particulier pour la composition FeCrMoNbB aux premiers stades de cristallisation. Les deux compositions présentaient d'excellentes propriétés de corrosion à l'état amorphe, avec un plateau de passivation s'étendant sur plus de 800 mV par rapport à l'OCP. Les données de Spectroscopie d'Impédance Électrochimique (EIS) ont été validées par le logiciel du modèle de mesure, et le modèle de Loi de Puissance a été appliqué pour interpréter les diagrammes, permettant le calcul de la résistivité à l'interface métal/film (ρ0) pour les deux alliages. La valeur de ρ0 était plus élevée pour l'alliage contenant du Nb, de l'ordre de 1013 Ωcm2. La spectroscopie photoélectronique des rayons X a été utilisée pour l'étude du film passif, et la couche passive compacte composée de Cr, Nb et Mo a été liée à la meilleure résistance à la corrosion de l'alliage contenant du Nb par rapport à celui contenant du Ni.Le comportement d'érosion-corrosion a été évalué en utilisant un disque des deux compositions généré à partir du processus de formage par pulvérisation de précurseurs commerciaux. Malgré des défauts de surface tels que des pores inhérents et des fissures résultant du processus d'application du revêtement, les deux revêtements présentaient une dureté supérieure de plus de trois fois à celle des aciers inoxydables conventionnellement utilisés. Le comportement d'érosion-corrosion des échantillons semblait être significativement influencé à la fois par la porosité et l'oxydation des particules. D'un point de vue électrochimique, tel que déterminé par des mesures de Potentiel de Circuit Ouvert (OCP), il était impossible de différencier entre les deux échantillons analysés.Dans une étape ultérieure, des revêtements ont été appliqués à l'aide de la technique de pulvérisation magnétron à courant continu, créant deux films minces amorphes compacts avec une grande dureté. Les résultats des revêtements étaient comparables à ceux obtenus à partir des échantillons de ruban. L'analyse du film passif via l'EIS et la Spectroscopie Photoélectronique des Rayons X (XPS) a permis de caractériser les films passifs décrits par le modèle Young pour l'alliage contenant du Nb et le modèle de Loi de Puissance pour l'alliage contenant du Ni. Les piqûres étaient absentes dans le cas du revêtement contenant du Nb, qui présentait des propriétés nettement supérieures par rapport au substrat.Cette enquête complète sur ces alliages à base de fer offre des informations précieuses sur leur potentiel pour les applications marines et industrielles, en abordant les défis de corrosion et d'érosion-corrosion. Ces matériaux ont démontré des performances exceptionnelles et une résistance à la corrosion, ce qui les positionne comme des alternatives viables aux aciers inoxydables conventionnels dans des conditions de fonctionnement difficiles
The quest for alternative materials to replace conventional stainless steel for marine applications has gained prominence recently. Among the emerging candidates, low-chromium Fe-based amorphous alloys have shown great promise, albeit needing a deeper understanding of their behavior. This research investigates the effect of devitrification on the corrosion and mechanical properties of two novel FeCrMoNbB and FeCrMoNiB pseudo high entropy amorphous compositions and their behavior when applied as coatings, shedding light on the critical role of passive film characterization.In the first stage, ribbons prepared through melt-spinning were subjected to annealing in an Ar protective atmosphere for simulation of crystallization. Corrosion tests were performed in a three-electrode cell, revealing the sensitivity of corrosion properties to the devitrification process, especially for the FeCrMoNbB composition in the first stages of crystallization. Both compositions presented outstanding corrosion properties in the amorphous state with a passivation plateau extending over 800mV relative to OCP. Electrochemical Spectroscopy Impedance (EIS) data was validated by Measurement Model software, and the Power Law model was applied to interpret the diagrams, allowing the calculation of resistivity at the metal/film interface (ρ0) for both alloys. The value of ρ0 was higher for the Nb-containing alloy, on the order of 1013 Ωcm2. X-ray photoelectron spectroscopy was applied for passive film study, and the compact passive layer composed of Cr, Nb, and Mo was linked to the superior corrosion resistance of the Nb-containing alloy compared to the Ni-containing one.Erosion-corrosion behavior was assessed using a disk of both compositions generated from commercial precursors' spray-forming process. Despite surface defects such as inherent pores and fissures resulting from the coating application process, both coatings exhibited hardness greater than three times that of conventionally used stainless steels. The erosion-corrosion behavior of the samples appeared to be significantly influenced by both porosity and particle oxidation. From an electrochemical perspective, as determined through Open Circuit Potential (OCP) measurements, it was impossible to differentiate between the two analyzed samples.In a subsequent step, coatings were applied using the DC magnetron sputtering technique, creating two compact, amorphous thin films with high hardness. The results from the coatings were comparable with those obtained from the ribbon samples. The passive film analysis via EIS and X-ray Photoelectron Spectroscopy XPS enabled the characterization of the passive films described by the Young model for the Nb-containing alloy and the Power Law model for the alloy containing Ni. Pitting was absent in the case of the Nb-containing coating, which exhibited markedly superior properties compared to the substrate.The comprehensive investigation of these Fe-based alloys offers valuable insights into their potential for marine and industrial applications, addressing corrosion and erosion-corrosion challenges. These materials have demonstrated outstanding performance and corrosion resistance, positioning them as viable alternatives to conventional stainless steels in harsh operating conditions

Grâce aux progrès des techniques de séquençage, les bases de données génomiques ont connu une croissance exponentielle depuis la fin des années 1990. Un grand nombre d'outils statistiques ont été développés à l'interface entre bioinformatique, apprentissage automatique et physique statistique, dans le but d'extraire de l'information de ce déluge de données. Plusieurs approches de physique statistique ont été récemment introduites dans le contexte précis de la modélisation de séquences de protéines, dont l'analyse en couplages directs. Cette méthode d'inférence statistique globale fondée sur le principe d'entropie maximale, s'est récemment montrée d'une efficacité redoutable pour prédire la structure tridimensionnelle de protéines, à partir de considérations purement statistiques.Dans cette thèse, nous présentons les méthodes d'inférence en question, et encouragés par leur succès, explorons d'autres domaines complexes dans lesquels elles pourraient être appliquées, comme la détection d'homologies. Contrairement à la prédiction des contacts entre résidus qui se limite à une information topologique sur le réseau d'interactions, ces nouveaux champs d'application exigent des considérations énergétiques globales et donc un modèle plus quantitatif et détaillé. À travers une étude approfondie sur des donnéesartificielles et biologiques, nous proposons une meilleure interpretation des paramètres centraux de ces méthodes d'inférence, jusqu'ici mal compris, notamment dans le cas d'un échantillonnage limité. Enfin, nous présentons une nouvelle procédure plus précise d'inférence de modèles génératifs, qui mène à des avancées importantes pour des données réelles en quantité limitée
Over the last decades, genomic databases have grown exponentially in size thanks to the constant progress of modern DNA sequencing. A large variety of statistical tools have been developed, at the interface between bioinformatics, machine learning, and statistical physics, to extract information from these ever increasing datasets. In the specific context of protein sequence data, several approaches have been recently introduced by statistical physicists, such as direct-coupling analysis, a global statistical inference method based on the maximum-entropy principle, that has proven to be extremely effective in predicting the three-dimensional structure of proteins from purely statistical considerations.In this dissertation, we review the relevant inference methods and, encouraged by their success, discuss their extension to other challenging fields, such as sequence folding prediction and homology detection. Contrary to residue-residue contact prediction, which relies on an intrinsically topological information about the network of interactions, these fields require global energetic considerations and therefore a more quantitative and detailed model. Through an extensive study on both artificial and biological data, we provide a better interpretation of the central inferred parameters, up to now poorly understood, especially in the limited sampling regime. Finally, we present a new and more precise procedure for the inference of generative models, which leads to further improvements on real, finitely sampled data

Framboids can be classified in terms of whether their constituent microcrystals are regularly arranged (ordered in a single domain), randomly arranged (disordered), or mixtures of both (partially ordered or multiple domains). The relative proportions of these three types are unknown, but there may be a tendency for the proportion of ordered framboids to increase with geologic age. In single domain microarchitectures the constituent microcrystals are uniform in both size and morphology and are arranged in a single array that is coherent through the framboid. Cubic close packing in these single domains explains most of the textures observed in sections of organized framboid, and this geometry maximizes entropy at high densities. Multiple domain microarchitectures can produce pseudo-icosahedral symmetries with 20 tetrahedral sub-units. Additionally, a number of other geometries have been reported, which do not readily fit either cubic close packed or icosahedral packing. Many apparently randomly ordered framboids show domains of organized microcrystals, which is consistent with the thesis that microcrystal self-organization evolves with time in the framboid. Similar organized domains are shown in some irregular masses of pyrite microcrystals, suggesting that self-organization and the development of the framboidal shape are largely independent processes.

Learning from label proportions (LLP) aims at learning an instance-level classifier with label proportions in grouped training data. Existing deep learning based LLP methods utilize end-to-end pipelines to obtain the proportional loss with Kullback-Leibler divergence between the bag-level prior and posterior class distributions. However, the unconstrained optimization on this objective can hardly reach a solution in accordance with the given proportions. Besides, concerning the probabilistic classifier, this strategy unavoidably results in high-entropy conditional class distributions at the instance level. These issues further degrade the performance of the instance-level classification. In this paper, we regard these problems as noisy pseudo labeling, and instead impose the strict proportion consistency on the classifier with a constrained optimization as a continuous training stage for existing LLP classifiers. In addition, we introduce the mixup strategy and symmetric cross-entropy to further reduce the label noise. Our framework is model-agnostic, and demonstrates compelling performance improvement in extensive experiments, when incorporated into other deep LLP models as a post-hoc phase.

The three dimensional, compressible, unsteady Navier-Stokes equations are solved to investigate the flow field of a centrifugal compressor with high compression ratio. Computational domain is consisted of an inlet bell and impeller with splitter blades followed by a two dimensional wedge vaned diffuser. The numerical method is validated by comparing the computational results with the experiments in terms of pressure ratio and compressor efficiency. The present study focuses on the unsteady pressure fluctuations and entropy production within the impeller and diffuser passages at the compressor design point. It is shown that the interaction between the impeller and diffuser blades leads to unsteadiness at the interface region and a pulsating behavior within the diffuser passages. Pressure waves with different convective velocities, generated by the impeller-diffuser interaction and pseudo-periodic unsteady separation bubbles, are captured in time/space domain along the diffuser blade surfaces. The pressure fluctuation spectra captured at the impeller-diffuser interface is evaluated to analyze the noise characteristics of the centrifugal compressor as a main source of blade passing frequency noise.

Existing adversarial domain adaptation methods mainly consider the marginal distribution and these methods may lead to either under transfer or negative transfer. To address this problem, we present a self-adaptive re-weighted adversarial domain adaptation approach, which tries to enhance domain alignment from the perspective of conditional distribution. In order to promote positive transfer and combat negative transfer, we reduce the weight of the adversarial loss for aligned features while increasing the adversarial force for those poorly aligned measured by the conditional entropy. Additionally, triplet loss leveraging source samples and pseudo-labeled target samples is employed on the confusing domain. Such metric loss ensures the distance of the intra-class sample pairs closer than the inter-class pairs to achieve the class-level alignment. In this way, the high accurate pseudolabeled target samples and semantic alignment can be captured simultaneously in the co-training process. Our method achieved low joint error of the ideal source and target hypothesis. The expected target error can then be upper bounded following Ben-David’s theorem. Empirical evidence demonstrates that the proposed model outperforms state of the arts on standard domain adaptation datasets.