On the differential benchmarking of promotional efficiency with machine learning modelling (II): Practical applications

The assessment of promotional sales with models constructed by machine learning techniques is arousing interest due, among other reasons, to the current economic situation leading to a more complex environment of simultaneous and concurrent promotional activities. An operative model diagnosis procedure was previously proposed in the companion paper, which can be readily used both for agile decision making on the architecture and implementation details of the machine learning algorithms, and for differential benchmarking among models. In this paper, a detailed example of model analysis is presented for two representative databases with different promotional behaviour, namely, a non-seasonal category (milk) and a heavily seasonal category (beer). The performance of four well-known machine learning techniques with increasing complexity is analyzed in detail here. In particular, k-Nearest Neighbours, General Regression Neural Networks, Multilayer Perceptron (MLP), and Support Vector Machines (SVM), are differentially compared. Present paper evaluates these techniques along the experiments described for both categories when applying the methodological findings obtained in the companion paper. We conclude that some elements included in the architecture are not essential for a good performance of the machine learning promotional models, such as the semiparametric nature of the kernel in SVM models, whereas other can be strongly dependent of the database, such as the convenience of multiple output models in MLP regression schemes. Additionally, the specificity of the behaviour of certain categories and product ranges determines the need to establish suitable and specific procedures for a better prediction and feature extraction.     

Influence of oxygen content on some properties of Nb3Sn tape used for superconducting transmission lines

The influence of the presence of oxygen in the Nb-Zr substrate upon some physical and mechanical properties of the Nb-ZrO₂ substrate and/or Nb₃Sn tape superconductor prepared from such a substrate has been studied. The results have shown that the Nb₃Sn grains are much finer in the case of Nb-ZrO₂, than those obtained on Nb-Zr substrate. As a result, both the electrical and mechanical properties have been substantially improved. The critical current at a field of 5 T increased by between 100 and 200 %, the alternating current losses decreased by about one order of magnitude, the tensile strength increased by about a factor of three and the microhardness value also increased by about a factor of three for oxygen contents in the range 0.03 to 0.70 wt%. The specific resistance of this substrate increased at the same time from 17 to 22 cm and the relative elongation decreased from 5 to 3 %.     

Support and precursor effects on the preparation of new heterogenized Pt/Sn catalysts for the selective hydroformylation of 1-pentene

New heterogenized Pt/Sn catalysts selective for the hydroformylation of 1-pentene have been synthesized. The complex cis-[PtCl₂(PPh₃)₂] and the SnCl₂.2H₂O or SnC₂O₄ precursors have been anchored on silica-, magnesia- and alumina-carriers. X-ray photoelectron spectroscopy was used to determine the surface composition and the nature of the anchored species. The hydroformylation activity was found to depend on the type of support and tin precursor used. Only the silica supported catalysts were active in the hydroformylation reaction. Samples prepared from SnCl₂-2H₂O were 200-fold more active than those prepared from SnC₂O₄. Selectivity ton-hexanal of the silica-supported catalyst prepared from SnCl₂-2H₂O was as high as 94.4% at 39.2% conversion of 1-pentene.     

Changes of copper location in CuY zeolites induced by preparation methods

The location of transition ions in copper- and copper-zinc-loaded Y type zeolites prepared by different procedures has been studied by temperature-programmed reduction, infrared spectroscopy of CO adsorbed on pretreated samples and X-ray photoelectron spectroscopy. Samples outgassed at 673 K showed Cu⁺ species due likely to reduction of Cu²⁺ ions under vacuum. Over exchanged CuY zeolites copper species in exchange sites were detected, while an impregnated sample exhibited bands of CO adsorbed on both Cu²⁺ and Cu⁺ ions developed at the surface of CuO crystals, and small proportions of Cu⁺ ions located in accessible exchange sites S_II and Sn_II. Similar findings were observed in Zn- and Cu-exchanged zeolites although the relative proportion of Cu in S_I positions was decreased due to competition between Cu²⁺ and Zn²⁺ ions. Samples reduced in hydrogen at 523 K showed the appearance of Cu⁰ species in impregnated samples, whereas Cu⁺ dominated in the exchanged counterparts. Reduction at 598 K led to substantial changes in Cu-exchanged samples in water. The proportion of Cu⁺ species decreased by reduction to Cu⁰ and simultaneously migration to Cu⁺ to S_II sites occurred. While Cu²⁺ or Cu⁺ were found on outgassed samples, only Cu⁰ and intrazeolite Cu⁺ were observed after H₂-reduction at 623 K. Changes in copper exposure as a function of sample pretreatments were also revealed by X-ray photoelectron spectroscopy.     

Leucinosis--a case of congenital metabolic disease in a child

Trasditional Chinese medicine is based on inductive-synthetic reasoning with regard to all life processes, in contrast to the causal-analytic orientation in our Western medicine. This philosophical approach is based on a system of comparisons (equivalents), referred to as Yin and Yang. Vital functions are seen as "energizing" processes, whereby the acupuncture points and the system of meridians are seen as a means of regulation for this "energy". A disturbed balance in this energy results in illness. The therapy according to the acupuncture method is to acheive a balance in the disturbed flow of energy by means of acupuncture points. By utilizing biophysical methods of measurement, it is possible to prove that a change in this energy takes place. The complex active principle which results in the effectiveness of acupuncture includes neurophysiological, biochemical and biophysical functional processes.     

Observation of a half-quantum-flux periodicity pattern in an rf SQUID with small critical currents

We have found experimentally that rf SQUID's working in a nonhysteretic mode with small critical currents give exactly at resonance and at very low rf bias a triangular pattern with a period equal to half of one elementary quantum of magnetic flux. We discuss a possible explanation of this peculiar result.     

A natural gas cash-out problem: A bilevel programming framework and a penalty function method

One of the many complex problems that arise from the transmission and marketing of natural gas is when a shipper draws a contract with a pipeline company to deliver a certain amount of gas among several points. What is actually delivered is often different from the amount that had been originally agreed upon. This phenomenon is called an imbalance. When an imbalance occurs, the pipeline penalizes the shipper by imposing a cash-out penalty policy. Since this penalty is a function of the operating daily imbalances, an important decision-making problem for the shippers is how to carry out their daily imbalances so as to minimize their incurred penalty. In this paper, we introduce the problem of minimizing the cash-out penalty costs from the point of view of a natural gas shipping party. We present a mixed integer bilevel linear programming model and discuss its underlying assumptions. To solve it efficiently, we reformulate it as a standard mathematical program and describe a penalty-function algorithm functions for its solution. The algorithm is well-founded and its convergence is proved. Results of numerical experiments support the algorithm’s robustness providing a valuable solution technique for this very important and complex problem in the natural gas market.     

Synthesis, structure, and field emission properties of sulfur-doped nanocrystalline diamond

The synthesis and properties of sulfur-doped nanocrystalline diamond films were investigated. The films were deposited by hot-filament chemical vapor deposition on Mo substrates using methane, hydrogen, and hydrogen disulfide. The nanocrystalline nature of the material arises from the induction of continuous secondary nucleation in the chemical vapor deposition environment. Complementary characterization tools were employed in order to obtain a comprehensive and coherent understanding of the correlations between the structural and electronic properties. In particular, sulfur-doped nanocrystalline diamond films show n-type Hall conductivity, enhanced field emission properties, and insensitivity to ion radiation. It was found that n-type doping of the tetragonally-bonded carbon matrix together with a nano network of trigonally-bonded carbon are crucial elements for enhanced field emission from nanocrystalline diamond. These conclusions and the corresponding supporting evidence are discussed.     

电力系统电流互感器饱和特性的柔性神经网络补偿法

电流互感器(CT)由于饱和使得副边电流变形,进而导致保护与测量应用中的许多问题。为此,提出一种补偿CT饱和特性的方法,以改善其测量性能。所提算法基于具有2个可变参数的sigmoid函数,构建了新型柔性神经网络,以估算CT励磁电流。实时地将估算所得励磁电流与扭曲的副边测量电流相加,即得补偿后原边电流。在学习过程中,所建补偿器的各柔性神经元柔性地改变其形状以适应各自的角色,高度柔性特点增强了网络学习能力,不但可减少网络节点数,而且可减少迭代学习时间。仿真研究中,应用一个900:5A的CT测试所提出的补偿器,测试时考虑了CT原边电流不同直流分量、CT剩磁大小与CT负载特性的影响。仿真结果验证了所提补偿方法的高精度,而且不受CT负载特性、CT剩磁情况及原边电流直流成分的影响。