Polymer 4D printing: Advanced shape-change and beyond

4D printing is an exciting branch of additive manufacturing. It relies on established 3D printing techniques to fabricate objects in much the same way. However, structures which fall into the 4D printed category have the ability to change with time, hence the “extra dimension.” The common perception of 4D printed objects is that of macroscopic single-material structures limited to point-to-point shape change only, in response to either heat or water. However, in the area of polymer 4D printing, recent advancements challenge this understanding. A host of new polymeric materials have been designed which display a variety of wonderful effects brought about by unconventional stimuli, and advanced additive manufacturing techniques have been developed to accommodate them. As a result, the horizons of polymer 4D printing have been broadened beyond what was initially thought possible. In this review, we showcase the many studies which evolve the very definition of polymer 4D printing, and reveal emerging areas of research integral to its advancement.     

Hyperparameter on-line learning of stochastic resonance based threshold networks

Aiming at training the feed-forward threshold neural network consisting of nondifferentiable activation functions, the approach of noise injection forms a stochastic resonance based threshold network that can be optimized by various gradient-based optimizers. The introduction of injected noise extends the noise level into the parameter space of the designed threshold network, but leads to a highly non-convex optimization landscape of the loss function. Thus, the hyperparameter on-line learning procedure with respective to network weights and noise levels becomes of challenge. It is shown that the Adam optimizer, as an adaptive variant of stochastic gradient descent, manifests its superior learning ability in training the stochastic resonance based threshold network effectively. Experimental results demonstrate the significant improvement of performance of the designed threshold network trained by the Adam optimizer for function approximation and image classification.     

Filters compatible with isomorphism testing

Like the lower central series of a nilpotent group, filters generalize the connection between nilpotent groups and graded Lie rings. However, unlike the case with the lower central series, the associated graded Lie ring may share few features with the original group: e.g. the associated Lie ring can be trivial or arbitrarily large. We determine properties of filters such that every isomorphism between groups is induced by an isomorphism between graded Lie rings.     

Periodicity of Two-Dimensional Bonding of Main Group Elements

In the periodic table the position of each atom follows the ‘aufbau’ principle of the individual electron shells. The resulting intrinsic periodicity of atomic properties determines the overall behavior of atoms in two-dimensional (2D) bonding and structure formation. Insight into the type and strength of bonding is the key in the discovery of innovative 2D materials. The primary features of 2D bonding and the ensuing monolayer structures of the main-group II–VI elements result from the number of valence electrons and the change of atom size, which determine the type of hybridization. The results reveal the tight connection between strength of bonding and bond length in 2D networks. The predictive power of the periodic table reveals general rules of bonding, the bonding-structure relationship, and allows an assessment of published data of 2D materials.     

基于GA_XGBoost模型的大学生科研能力预测问题研究

科学评价大学生科研创新能力对我国科研水平的提高具有重要意义.采用机器学习模型来预测大学生科研能力可以起到良好的效果,提出一种GAXGBoost模型来实现对大学生的科研能力预测.此模型是以Xgboost算法为基础,然后充分利用遗传算法的全局搜索能力自动搜索Xgboost最优超参数,避免了人为经验调参不准确的缺陷,最后采用精英选择策略以此确保每一轮都是最佳的进化结果.通过分析表明,所采用的GAXGBoost模型在大学生科研能力预测的结果中具有很高的精度,将此模型与Logistic Regression、Random Forest、SVM等模型进行对比,GAXGBoost模型的预测精度最高.     

Calcium-based coordination polymers from a solvothermal synthesis of HKUST-1 in 3D printed autoclaves

A mixed-metal 1D coordination polymer [CaCu(HBTC)₂(H2O)₈]_n (where H₃BTC – benzene-1,3,5-tric arboxylic acid) was obtained in a solvothermal synthesis of a well-known copper-containing metal–organic framework [Cu₃(BTC)₂(H₂O)₃]_n (HKUST-1) in autoclaves 3D-printed from commercial polypropylene. This material was a source of calcium ions, apparently, leaking from a colorant (calcium carbonate) promoted by glacial acetic acid as a modulator used to produce large single crystals of HKUST-1. This finding was confirmed by elemental analysis and a model experiment that resulted in a new calcium-based 1D coordination polymer [Ca(H₂BTC)₂(H₂O)₅]_n under the same solvothermal conditions with no copper or calcium salts put into a 3D-printed autoclave.     

A branch-and-bound method for the minimum k−enclosing ball problem

The minimum k-enclosing ball problem seeks the ball with smallest radius that contains at least k of m given points. This problem is NP-hard. We present a branch-and-bound algorithm on the tree of the subsets of k points to solve this problem. Our method is able to solve the problem exactly in a short amount of time for small and medium sized datasets.