Explaining prediction models and individual predictions with feature contributions

We present a sensitivity analysis-based method for explaining prediction models that can be applied to any type of classification or regression model. Its advantage over existing general methods is that all subsets of input features are perturbed, so interactions and redundancies between features are taken into account. Furthermore, when explaining an additive model, the method is equivalent to commonly used additive model-specific methods. We illustrate the method’s usefulness with examples from artificial and real-world data sets and an empirical analysis of running times. Results from a controlled experiment with 122 participants suggest that the method’s explanations improved the participants’ understanding of the model.     

Connective stability of large-scale stochastic systems†

In this work, we will introduce the notion of connective stability in the mean for large-scale stochastic systems under structural perturbations. Sufficient conditions for this kind of stability are obtained by the decomposition-aggregation method in the context of vector Liapuiiov functions.     

Control of mineral wool thickness using predictive functional control

The production process of mineral wool is affected by several constantly changing factors. The ingredients for the mineral wool are melted in a furnace. The molten mineral charge exits the bottom of the furnace in a water-cooled trough and falls into a fiberization device (the centrifuge). The centrifuge forms the fibers. At this stage binders are injected to bind the fibers together. To ensure the quality of the end product (the consistent thickness) the flow of the bounded fibers must be as constant as possible. One way to ensure that is to control the speed of the conveyor belt that transports the bounded fibers from the centrifuge to the curing process. Predictive functional controller and PID controller are considered to replace an existing algorithm. Both can easily replace an existing one as they do not require any new sensor installation. All three algorithms are presented and tested on a developed plant model. The study showed that the predictive control gives better results than the existing and PID controller.     

Full theory of speckle displacement and decorrelation in the image field by wave and geometrical descriptions and its application in mechanics

Abstract

The paper describes the possibility of using the speckle pattern decorrelation for determination of small deformation tensor components of an elementary object surface area in an optical image field. The relationship between the small-deformation tensor and the speckle field displacement is analysed in detail. The studied problem is presented from the approximation viewpoint of both wave and geometrical optics.     

Bread and durum wheat compared for antioxidants contents,and lipoxygenase and peroxidase activities

Potential beneficial components, including proteins, total phenolics, total flavonoids, carotenoids, tocopherols, and DPPH radical scavenging activity, were investigated in wholemeal of ten bread (T. aestivum L.) and ten durum (T. durum Desf.) novel wheat genotypes. In addition, the activity rate of lipoxygenase (LOX) and peroxidase (POD) enzymes implicated in the antioxidant metabolism was determined. The protein contnet and the antioxidant properties varied according to the two different wheat species, as well as, between the different bread and durum wheat genotypes themselves. The results indicated significant differences in proteins and antioxidant compounds between bread and durum wheat. Higher total proteins, wet gluten and antioxidants contents, combined with lower LOX and POD activities, point to a higher nutritive value of durum wheat than bread wheat.     

Multiscale approach to the morphology,structure, and segmental dynamics of complex degradable aliphatic polyurethanes

A multiscale approach spanning from the segmental (subnanometer) up to micrometer level was applied for detailed study of the self‐assembly of aliphatic block polyurethane (PU) elastomers. To understand the principles of the self‐organization of hard and soft segments in the complex multi‐component systems, several two‐component model PU samples, that is, the products of 1,6‐diisocyanatohexane (HDI) with three diols differing in the length and constitution were also prepared, characterized, and investigated: (i) polycarbonate‐based macrodiol (MD), (ii) biodegradable oligomeric diol (DL‐L; product of butane‐1,4‐diol and D,L‐lactide), and (iii) butane‐1,4‐diol (BD). The study (particularly ¹³C‐¹H PILGRIM NMR spectra) reveals complex internal organization and interesting (application appealing) behavior of multi‐component PUs. Hard segments (HDI+BD products) feature self‐assembled and significantly folded chain conformations with interdomain spacing 15–22 nm (small‐angle X‐ray scattering analysis). The small domains are hierarchically assembled in various structural formations of µm size (spherulites) depending on PU composition, as detected by transmission electron microscopy and atomic force microscopy. © 2014 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015 , 132, 41590.     

On the solid state heat transport phenomena measurement

The paper deals with theoretical and experimental aspects of lumped capacitance model (LCM) application for the study of heat transport in different materials. Patented construction of the measuring chamber together with special software the fundamental features of which are presented here allows evaluating thermal conductivity k, specific heat capacity c_p and thermal diffusivity α. Obtained results are in relatively very good agreement with those obtained from independent measurement or table values.     

Influence of nitriding on corrosion resistance of martensitic X17CrNi16‐2 stainless steel

Nitriding increases surface hardness and improves wear resistance of stainless steels. However, nitriding can sometimes reduce their corrosion resistance. In this paper, the influence of nitriding on the corrosion resistance of martensitic stainless steel was investigated. Plasma nitriding at 440 °C and 525 °C and salt bath nitrocarburizing were carried out on X17CrNi16‐2 stainless steel. Microhardness profiles of the obtained nitrided layers were examined. Phase composition analysis and quantitative depth profile analysis of the nitrided layers were preformed by X‐ray diffraction (XRD) and glow‐discharge optical emission spectrometry (GD‐OES), respectively. Corrosion behaviour was evaluated by immersion test in 1% HCl, salt spray test in 5% NaCl and electrochemical corrosion tests in 3.5% NaCl aqueous solution. Results show that salt bath nitrocarburizing, as well as plasma nitriding at low temperature, increased microhardness without significantly reducing corrosion resistance. Plasma nitriding at a higher temperature increased the corrosion tendency of the X17CrNi16‐2 steel.