An index structure supporting rule activation in pervasive applications

World Wide Web - Rule mechanism has been widely used in many areas, such as databases, artificial intelligent and pervasive computing. In a rule mechanism, rule activation decides which rules are...     

Synthesis of Ionic Ultramicroporous Polymers for Selective Separation of Acetylene from Ethylene

The design of highly stable and efficient porous materials is essential for developing breakthrough hydrocarbon separation methods based on physisorption to replace currently used energy-intensive distillation/absorption technologies. Efforts to develop advanced porous materials such as zeolites, coordination frameworks, and organic polymers have met with limited success. Here, a new class of ionic ultramicroporous polymers (IUPs) with high-density inorganic anions and narrowly distributed ultramicroporosity is reported, which are synthesized by a facile free-radical polymerization using branched and amphiphilic ionic compounds as reactive monomers. A covalent and ionic dual-crosslinking strategy is proposed to manipulate the pore structure of amorphous polymers at the ultramicroporous scale. The IUPs exhibit exceptional selectivity (286.1–474.4) for separating acetylene from ethylene along with high thermal and water stability, collaboratively demonstrated by gas adsorption isotherms and experimental breakthrough curves. Modeling studies unveil the specific binding sites for acetylene capture as well as the interconnected ultramicroporosity for size sieving. The porosity-engineering protocol used in this work can also be extended to the design of other ultramicroporous materials for the challenging separation of other key gas constituents.     

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.     

Design of delayed reconfigurable manufacturing system based on part family grouping and machine selection

To improve the convertibility of reconfigurable manufacturing system (RMS), the concept of delayed reconfigurable manufacturing system (D-RMS) was proposed. RMS and D-RMS are both constructed around part family. However, D-RMS may suffer from ultra-long system problem with unacceptable idle machines using generic RMS part families. Besides, considering the complex basic system structure of D-RMS, machine selection of D-RMS should be addressed, including dedicated machine, flexible machine, and reconfigurable machine. Therefore, a system design method for D-RMS based on part family grouping and machine selection is proposed. Firstly, a part family grouping method is proposed for D-RMS that groups the parts with more former common operations into the same part family. The concept of longest relative position common operation subsequence (LPCS) is proposed. The similarity coefficient among the parts is calculated based on LPCS. The reciprocal value of the operation position of LPCS is adopted as the characteristic value. The average linkage clustering (ALC) algorithm is used to cluster the parts. Secondly, a machine selection method is proposed to complete the system design of D-RMS, including machine selection rules and the dividing point decision model. Finally, a case study is given to implement and verify the proposed system design method for D-RMS. The results show that the proposed system design method is effective, which can group parts with more former common operations into the same part family and select appropriate machine types.