Search

Slip transfer and blocking across grain boundaries was studied in a Ti foil with a strong rolling texture deformed in tension. Prior to deformation, the shape of the grains and the orientation of the grain boundaries were quantified through laboratory scale diffraction contrast tomography (LabDCT). Mechanical deformation led to the activation of prismatic slip, and slip transfer/blocking was assessed in > 300 grain boundaries by means of slip trace analysis and electron backscatter diffraction. A categorical model was employed to accurately assess slip transfer, and the "F1 score" of various slip transfer criteria proposed in the literature was evaluated for the first time from 3D grain boundary information. Remarkably, for the prismatic-dominated slip transfer in the current Ti sample, the results show that the best predictions of slip transfer/blocking are provided by the angle \k{appa}, which is directly related to the residual Burgers vector, and by the Luster-Morris parameter m'. In contrast, metrics based on the twist angle {\theta} and on the LRB criterion were not able to predict accurately slip transfer/blocking. Thus, the extensive analysis of the 3D grain boundary data and the novel application of LabDCT was able to help clarify the role of grain boundary orientation on the mechanisms of plastic deformation in polycrystals with strong prismatic-dominated slip.

The fracture toughness of Al/SiC nanolaminates with different layer thicknesses (in the range 10 to 100 nm) was measured by means of micropillar splitting and bending of a notched beam. The crack plane was perpendicular to the layers in the former while notched beams with the notch parallel and perpendicular to the layers were milled in the latter. It was found that crack propagation parallel to the layers took place along the metal-ceramic interfaces and the toughness increased with the layer thickness due to the contribution of the plastic deformation of the Al layers. Crack propagation perpendicular to the layers showed evidence of crack deflection/arrest at the interface. The toughness in this orientation increased as the layer thickness decreased due to the higher density of interfaces except for the nanolaminates with 10 nm layer thickness. In the latter case, crack propagation took place along the weak columnar grain boundaries, leading to a marked reduction in toughness.

The spread of non-indigenous species (NIS) through the global shipping network (GSN) has enormous ecological and economic cost throughout the world. Previous attempts at quantifying NIS invasions have mostly taken "bottom-up" approaches that eventually require the use of multiple simplifying assumptions due to insufficiency and/or uncertainty of available data. By modeling implicit species exchanges via a graph abstraction that we refer to as the Species Flow Network (SFN), a different approach that exploits the power of network science methods in extracting knowledge from largely incomplete data is presented. Here, coarse-grained species flow dynamics are studied via a graph clustering approach that decomposes the SFN to clusters of ports and inter-cluster connections. With this decomposition of ports in place, NIS flow among clusters can be very efficiently reduced by enforcing NIS management on a few chosen inter-cluster connections. Furthermore, efficient NIS management strategy for species exchanges within a cluster (often difficult due higher rate of travel and pathways) are then derived in conjunction with ecological and environmental aspects that govern the species establishment. The benefits of the presented approach include robustness to data uncertainties, implicit incorporation of "stepping-stone" spread of invasive species, and decoupling of species spread and establishment risk estimation. Our analysis of a multi-year (1997--2006) GSN dataset using the presented approach shows the existence of a few large clusters of ports with higher intra-cluster species flow that are fairly stable over time. Furthermore, detailed investigations were carried out on vessel types, ports, and inter-cluster connections. Finally, our observations are discussed in the context of known NIS invasions and future research directions are also presented.

There has been increased interest in devising learning techniques that combine unlabeled data with labeled data ? i.e. semi-supervised learning. However, to the best of our knowledge, no study has been performed across various techniques and different types and amounts of labeled and unlabeled data. Moreover, most of the published work on semi-supervised learning techniques assumes that the labeled and unlabeled data come from the same distribution. It is possible for the labeling process to be associated with a selection bias such that the distributions of data points in the labeled and unlabeled sets are different. Not correcting for such bias can result in biased function approximation with potentially poor performance. In this paper, we present an empirical study of various semi-supervised learning techniques on a variety of datasets. We attempt to answer various questions such as the effect of independence or relevance amongst features, the effect of the size of the labeled and unlabeled sets and the effect of noise. We also investigate the impact of sample-selection bias on the semi-supervised learning techniques under study and implement a bivariate probit technique particularly designed to correct for such bias.

An approach to the construction of classifiers from imbalanced datasets is described. A dataset is imbalanced if the classification categories are not approximately equally represented. Often real-world data sets are predominately composed of "normal" examples with only a small percentage of "abnormal" or "interesting" examples. It is also the case that the cost of misclassifying an abnormal (interesting) example as a normal example is often much higher than the cost of the reverse error. Under-sampling of the majority (normal) class has been proposed as a good means of increasing the sensitivity of a classifier to the minority class. This paper shows that a combination of our method of over-sampling the minority (abnormal) class and under-sampling the majority (normal) class can achieve better classifier performance (in ROC space) than only under-sampling the majority class. This paper also shows that a combination of our method of over-sampling the minority class and under-sampling the majority class can achieve better classifier performance (in ROC space) than varying the loss ratios in Ripper or class priors in Naive Bayes. Our method of over-sampling the minority class involves creating synthetic minority class examples. Experiments are performed using C4.5, Ripper and a Naive Bayes classifier. The method is evaluated using the area under the Receiver Operating Characteristic curve (AUC) and the ROC convex hull strategy.