A self-adaptive synthetic over-sampling technique for imbalanced classification

Traditionally, in supervised machine learning, (a significant) part of the available data (usually 50%-80%) is used for training and the rest—for validation. In many problems, however, the data are highly imbalanced in regard to different classes or does not have good coverage of the feasible data space which, in turn, creates problems in validation and usage phase. In this paper, we propose a technique for synthesizing feasible and likely data to help balance the classes as well as to boost the performance in terms of confusion matrix as well as overall. The idea, in a nutshell, is to synthesize data samples in close vicinity to the actual data samples specifically for the less represented (minority) classes. This has also implications to the so-called fairness of machine learning. In this paper, we propose a specific method for synthesizing data in a way to balance the classes and boost the performance, especially of the minority classes. It is generic and can be applied to different base algorithms, for example, support vector machines, k-nearest neighbour classifiers deep neural, rule-based classifiers, decision trees, and so forth. The results demonstrated that (a) a significantly more balanced (and fair) classification results can be achieved and (b) that the overall performance as well as the performance per class measured by confusion matrix can be boosted. In addition, this approach can be very valuable for the cases when the number of actual available labelled data is small which itself is one of the problems of the contemporary machine learning.     

Interactive design exploration for constrained meshes

In architectural design, surface shapes are commonly subject to geometric constraints imposed by material, fabrication or assembly. Rationalization algorithms can convert a freeform design into a form feasible for production, but often require design modifications that might not comply with the design intent. In addition, they only offer limited support for exploring alternative feasible shapes, due to the high complexity of the optimization algorithm.We address these shortcomings and present a computational framework for interactive shape exploration of discrete geometric structures in the context of freeform architectural design. Our method is formulated as a mesh optimization subject to shape constraints. Our formulation can enforce soft constraints and hard constraints at the same time, and handles equality constraints and inequality constraints in a unified way. We propose a novel numerical solver that splits the optimization into a sequence of simple subproblems that can be solved efficiently and accurately.Based on this algorithm, we develop a system that allows the user to explore designs satisfying geometric constraints. Our system offers full control over the exploration process, by providing direct access to the specification of the design space. At the same time, the complexity of the underlying optimization is hidden from the user, who communicates with the system through intuitive interfaces.     

Platinum nanoparticles decorated on graphitic carbon nitride-ZIF-67 composite support: An electrocatalyst for the oxidation of butanol in fuel cell applications

In the present study, we report an eco-friendly and simple route to design and synthesize novel nanocomposite catalyst based on platinum nanoparticles anchored on binary support of graphitic carbon nitride (g-C₃N₄) and cobalt-metal-organic framework (ZIF-67). For this purpose, ZIF-67 was prepared by precipitation method and g-C₃N₄ was prepared through thermal polymerization method. Later, ZIF-67 and g-C₃N₄ were hybridized through sonication to get homogeneous g–C₃N₄–ZIF-67 nanocomposite support material. Platinum nanoparticles (PtNPs) were uniformly deposited on g–C₃N₄–ZIF-67 by an electrochemical method. The as-developed nanocatalyst was characterized by morphological, structural and electrochemical techniques. The electrocatalytic activity of PtNPs@g–C₃N₄–ZIF-67 nanocatalyst towards butanol oxidation was evaluated via CV, CA, LSV and EIS in an alkaline medium. Results revealed that the proposed catalyst showed greatly enhanced electrooxidation of butanol in terms of high magnificent current density, lower oxidation potential, excellent long-term stability, large surface area, low charge transfer resistance and less toxic ability. Enhanced catalytic performance of the proposed catalyst could be ascribed to the synergistic effect of g–C₃N₄–ZIF-67 nanocomposite and PtNPs. The PtNPs@g–C₃N₄–ZIF-67 catalyst holds promising potential applications to be used as an anodic electrocatalyst for the development of high-performance alkaline fuel cells.     

Output-relevant Variational autoencoder for Just-in-time soft sensor modeling with missing data

Main challenges for developing data-based models lie in the existence of high-dimensional and possibly missing observations that exist in stored data from industry process. Variational autoencoder (VAE) as one of the deep learning methods has been applied for extracting useful information or features from high-dimensional dataset. Considering that existing VAE is unsupervised, an output-relevant VAE is proposed for extracting output-relevant features in this work. By using correlation between process variables, different weight is correspondingly assigned to each input variable. With symmetric Kullback–Leibler (SKL) divergence, the similarity is evaluated between the stored samples and a query sample. According to the values of the SKL divergence, data relevant for modeling are selected. Subsequently, Gaussian process regression (GPR) is utilized to establish a model between the input and the corresponding output at the query sample. In addition, owing to the common existence of missing data in output data set, the parameters and missing data in the GPR are estimated simultaneously. A practical debutanizer industrial process is utilized to illustrate the effectiveness of the proposed method.     

不同硬度PTMG型CPUE与硫化橡胶的性能对比研究

通过改变硬段含量、扩链系数等制备了几种不同硬度的四氢呋喃均聚醚(PTMG)浇注型聚氨酯弹性体(CPUE),讨论了PTMG型CPUE与硫化橡胶(VR)的物理机械性能、拉伸应力应变行为以及动态力学性能的区别。结果表明,随CPUE硬度的提高,材料的刚性增强,力学性能是低硬度VR的数倍。通过对CPUE硬度的调节,可以使之兼具聚氨酯和橡胶的优良性能。在玻璃态温度下CPUE的储能模量随硬度的减小而降低,但都高于VR,而且CPUE的硬度越大,二者相差越大,温度越低,相差也越大;在室温以上温度范围内,CPUE具有比VR更小的力学损耗。     

Temperature adaptability of the soluble lead flow battery using different solutions of fluoborate,perchlorate, methanesulfonate and trifluoromethanesulfonate

The evaluation of cell's weatherability is of practical interest. To further improve the soluble lead flow battery's weatherability, physiochemical properties of electrolytes containing fluoborate, perchlorate, methanesulfonate and trifluoromethanesulfonate are investigated from ?60 to 50 °C. Activities of CF₃SO₃H and HClO₄ are poor in trifluoromethanesulfonate and perchlorate solutions due to common anion effect. The solubility of lead salt can be improved by increasing temperature, but worsened by increasing acid's content. With the temperature increasing, the conductivity is enhanced, and the viscosity is lowered for four solutions. The same results have been found by increasing acid's content except for CF₃SO₃H. The high energy efficiency can be achieved for cells over ?40–0 °C using fluoborate and perchlorate solutions, 73.2% at ?40 °C and 78.1% at ?30 °C respectively. Over the temperature range of 20–50 °C, the cells with methanesulfonate and trifluoromethanesulfonate solutions have good performance, 77.4% and 73.7% at 50 °C respectively.     

Reliability of complex chemical engineering processes

This paper presents a stochastic performance modelling approach that can be used to optimise design and operational reliability of complex chemical engineering processes. The framework can be applied to processes comprising multiple units, including the cases where closed form process performance functions are unavailable or difficult to derive from first principles, which is often the case in practice. An interface that facilitates automated two-way communication between Matlab^® and process simulation environment is used to generate large process responses. The resulting constrained optimisation problem is solved using both Monte Carlo Simulation (MCS) and First Order Reliability Method (FORM); providing a wide range of stochastic process performance measures. Adding such capabilities to traditional deterministic process simulators provides a more informed basis for selecting optimum design factors; giving a simple way of enhancing overall process reliability and cost-efficiency. Two case study systems are considered to highlight the applicability and benefits of the approach.     

Synthesis and Characterization of Co Nanocomposite Based on Poly(benzimidazole-amide) Matrix and Their Behavior as Catalyst in Oxidation Reaction

2,6-Bis(5-amino-1H-benzimidazol-2-yl)pyridine was prepared and characterized by Fourier transform infrared spectroscopy, elemental analysis, ¹H-NMR, and ¹³C-NMR spectroscopic methods. Then a new poly(benzimidazole-amide) was synthesized by polymerization of the corresponding diamine and isophthalic acid. The obtained poly(benzimidazole-amide) exhibited good yield and high thermal stability. Due to the existence of benzimidazole moieties in polymer’s structure, it has the tendency to form complexes with metal ions. So, a new poly(benzimidazole-amide)/Co nanocomposite was prepared. Morphological studies revealed that metal nanoparticles were dispersed in the polymer matrix without any aggregation. poly(benzimidazole-amide)/Co nanocomposite was used as a catalyst in the oxidation of ethyl benzene to acetophenone with tert-butyl hydroperoxide.     

基于GFU和分层LSTM的组群行为识别研究方法

提出一种以"关键人物"为核心，使用门控融合单元（GFU，Gated Fusion Unit）进行特征融合的组群行为识别框架，旨在解决两个问题：①组群行为信息冗余，重点关注关键人物行为特征，忽略无关人员对组群行为的影响；②组群内部交互行为复杂，使用GFU有效融合以关键人物为核心的交互特征，再通过LSTM时序建模成为表征能力更强的组群特征.最终，通过softmax分类器进行组群行为类别分类.该算法在排球数据集上取得了86.7%的平均识别率.