No abstract

In this paper, ammonium polyphosphate(APP)/expandable graphite(EG)/thermoplastic polyurethane (TPU) composites were prepared by microlayer coextrusion technology, APP and EG fillers had good synergistic flame retardancy and excellent dispersion in TPU matrix, which greatly improved the flame retardancy and mechanical properties of multilayer composites. The dispersion of APP and EG in TPU was characterized by using SEM, the flame retardancy of composites was characterized by using UL94 and LOI, the thermal stability of composites was characterized by using TGA and DTG, and tensile test was used to characterized the mechanical properties of composites. SEM showed that the microlayer coextrusion technology significantly improved the dispersion of APP and EG in TPU. As showed by the experimental results, the vertical combustion level of ordinary blended composites reached V‐2 after adding only one kind of filler either APP or EG, and the vertical combustion level of ordinary blended composites reached V‐0 with APP and EG applied together, while the vertical combustion level of microlayer coextruded composites all reached V‐0 when the total addition of APP and EG was 15%. In particular, the LOI value of microlayer coextruded composites was 30.9%, while the LOI value of ordinary blended composites only was 27.9% when APP: EG = 1: 1. At this time, the flame retardancy level of APP/EG/TPU composites was the best. In addition, the thermal stability and mechanical properties of microlayer coextruded composites were far superior to ordinary blended composites. In conclusion, the synergistic flame retardancy of APP and EG fillers and the dispersion of APP and EG fillers in TPU matrix played a significant role in enhancing flame retardancy and mechanical properties.

Consider the following class of learning schemes:where x i ∈ R p and y i ∈ R denote the i th feature and response variable respectively. Let and R be the convex loss function and regularizer, β denote the unknown weights, and λ be a regularization parameter. C ⊂ R p is a closed convex set. Finding the optimal choice of λ is a challenging problem in high-dimensional regimes where both n and p are large. We propose three frameworks to obtain a computationally efficient approximation of the leave-one-out cross validation (LOOCV) risk for nonsmooth losses and regularizers. Our three frameworks are based on the primal, dual, and proximal formulations of (1). Each framework shows its strength in certain types of problems. We prove the equivalence of the three approaches under smoothness conditions. This equivalence enables us to justify the accuracy of the three methods under such conditions. We use our approaches to obtain a risk estimate for several standard problems, including generalized LASSO, nuclear norm regularization, and support vector machines. We empirically demonstrate the effectiveness of our results for non-differentiable cases.

FAdV-4, as the main cause of HHS, has quickly spread all over the world in recent years, seriously threatening the poultry industry. The aim of this study was to identify the important host proteins that have the potential to regulate the life cycle of FAdV-4.

With the rapid growth of the global economy, the automatic recognition of financial bills becomes the primary way to reduce the burden of the traditional manual approach for bill recognition and classification. However, most automatic recognition methods cannot effectively recognize the handwritten characters on financial bills, especially when the bills come from different financial companies. To solve the problem, this paper fully explores the bill system in banks and the operations of bill number recognition, and then develops a hybrid classifier based on deep convolutional neural network (DCNN) and support vector machine (SVM), with the aim to recognize the handwritten numbers on financial bills in different domains. The DCNN with different channels was adopted to effectively mine the local handwritten numbers on financial bills from varied sources. Then, the extracted information was fed to the SVM to realize accurate classification of numbers. Our method makes full use of the distribution difference between information in different fields, and adapts to different fields based on the parameter sharing mechanism. Experimental results show that our method can recognize the handwritten numbers on financial bills more accurately (>3%) than benchmark methods.