117,529 research outputs found
Ceramics: Chemical and Petrographic Analysis
No full textAncient ceramics, the most abundant findings in archaeological sites since prehistory, are mobile objects that could have been exchanged/traded far away from their production center. Their provenance and production technology represent, therefore, an important issue that can be studied through petrographic and chemical analyses to discuss cultural differences, exchange routes, and technological developments. Using petrography, the occurrence (especially in coarse ware) of mineral/ rock markers allows to link their production with specific geological regions, while microstructural and textural features of the ceramic body can supply information on their production technology in terms of forming techniques and paste recipes. When dealing especially with fine ware, the bulk geochemical composition can identify specific production centers through comparison with ceramic reference groups, and reveal possible preparation procedures of the base clay, such as depuration and mixing. Ceramic coatings such as slips, paints and glazes can be analyzed under a microscope and their elemental composition can be further assessed by micro-chemical and isotopic analyses
I laterizi iscritti di epoca romana rinvenuti nel crollo del campanile di San Marco. Nuovi dati da vecchi scavi.
No full textPottery production at the mesolithic site of Kabbashi Haitah (central Sudan) : an integrated morphological, petrographic and mineralogical analysis
No full textThis paper presents the results of an interdisciplinary archeometric study on Early Mesolithic pottery from the prehistoric site of Kabbashi Haitah, located 35 km north of Khartoum (central Sudan), along the Nile Valley. A large set of potsherds, selected after a preliminary macroscopic analysis of 1075 fragments representing the various vessels (mainly plain and globular in shape, with various rim diameter), macrofabrics and decoration types (either with or without incised or stamped decorations, i.e. incisedwavy line and rocker stamp) was analysed to define the type of the raw materials used and their manufacturing technology. The mineralogical and petrographic features, determined by optical microscopy and X-ray powder diffraction, indicate that the pottery was produced using an illitic clay tempered with quartz and/or K-feldspar derived from granite/syenite grinding, and fired in the temperature range between 750 and 900 degrees C
A Renormalization Group approach to dynamical properties of hierarchical networks
No full textDiffusion processes in the presence of hierarchical distributions of transition rates or waiting times are investigated by Renormalization Group (RG) techniques. Diffusion on one-dimensional chains, loop-less fractals and fully ultrametric spaces are considered. RG techniques are shown to be most natural and powerful to apply when infinitely many time scales are simultaneously involved in a problem. Generalizations and extensions of existing models and results are easily accomplished in the RG context. Wherever possible, heuristic scaling arguments are also presented in order to give an easier physical interpretation of the analytical results.
Two relevant applications of ultradiffusion models are reviewed in detail. One of them concerns breakdown of dynamic scaling in a one-dimensional hierarchical Glauber chain. The other one is in the context of tethered random surface models
Hierarchical Random Surfaces
No full textTwo-dimensional random surfaces are constructed by the mapping in d-dimensional space of a triangular network (of linear size L) with hierarchical bond structure. A relation between static properties of such surfaces and the resistance exponent, ζ, of two-dimensional inhomogeneous structures allows us to show that free surfaces have an average radius of gyration ξ∼Lζ/2 for L→∞. When a self-avoidance constraint is imposed, a Flory argument gives ξ∼L2/D with D=(2+d)/(2+(1/2ζ), and an upper critical dimension dc=8/ζ. Specific examples are discussed where exact, nontrivial values of ξ can be predicted by renormalization and scaling arguments. Universal, as well as nonuniversal, asymptotic regimes occur
Foreword: Multidisciplinary study of the Sarno Baths in Pompeii (Naples, Italy): Preface
No full textDo veins and stomata patterns in angiosperms obey general rules?
No full textPlants have evolved a huge variety of leaf forms, vein structures, stomata dimensions and behavior. Here we investigate whether general principles might be evoked to explain the distribution of leaf veins and stomata across species. We analyzed 20 angiosperms grown in different environments (alpine, middle latitude, tropics). We took 5-15 mm2 samples from the proximal, middle and distal part of a mature leaf of each species. Stomata imprints were collected using acrylic varnish on transparent tape; veins were highlighted by chemical treatment using NaOH. With a semi-automatic GIS-based procedure we identified the smallest leaf area completely enclosed by veins as a loop and we measured area, contour, stomata number, pore lengths for each loop. About 60-700 loops were measured for each species. Stomata density was species-specific (ranging 50-350 pore/mm2) and seemed not to change with position on the leaf nor with the order of surrounding veins. Notably, in each loop the number of stomata scaled almost isometrically with loop area (exponents ranging 0.9-1.0) (i.e. Ns~A1) and loop contour (0.9-1.2) (i.e. Ns~L1) in all measured species; moreover the scaling coefficients appeared related to pore length. This would suggest that a simple scale-free pattern has evolved regardless of species or environment
FedZeN: Quadratic Convergence in Zeroth-Order Federated Learning via Incremental Hessian Estimation
No full textFederated learning is a distributed learning framework that allows a set of clients to collaboratively train a model under the orchestration of a central server, without sharing raw data samples. Although in many practical scenarios the derivatives of the objective function are not available, only few works have considered the federated zeroth-order setting, in which functions can only be accessed through a budgeted number of point evaluations. In this work we focus on convex optimization and design the first federated zeroth-order algorithm to estimate the curvature of the global objective, with the purpose of achieving superlinear convergence. We take an incremental Hessian estimator whose error norm converges linearly in expectation, and we adapt it to the federated zeroth-order setting, sampling the random search directions from the Stiefel manifold for improved performance. Both the gradient and Hessian estimators are built at the central server in a communication-efficient and privacy-preserving way by leveraging synchronized pseudo-random number generators. We provide a theoretical analysis of our algorithm, named FedZeN, proving local quadratic convergence with high probability and global linear convergence up to zeroth-order precision. Numerical simulations confirm the superlinear convergence rate and show that our algorithm outperforms the federated zeroth-order methods available in the literature
- …
