ENDER: a statistical framework for boosting decision rules

Induction of decision rules plays an important role in machine learning. The main advantage of decision rules is their simplicity and human-interpretable form. Moreover, they are capable of modeling complex interactions between attributes. In this paper, we thoroughly analyze a learning algorithm, called ENDER, which constructs an ensemble of decision rules. This algorithm is tailored for regression and binary classification problems. It uses the boosting approach for learning, which can be treated as generalization of sequential covering. Each new rule is fitted by focusing on examples which were the hardest to classify correctly by the rules already present in the ensemble. We consider different loss functions and minimization techniques often encountered in the boosting framework. The minimization techniques are used to derive impurity measures which control construction of single decision rules. Properties of four different impurity measures are analyzed with respect to the trade-off between misclassification (discrimination) and coverage (completeness) of the rule. Moreover, we consider regularization consisting of shrinking and sampling. Finally, we compare the ENDER algorithm with other well-known decision rule learners such as SLIPPER, LRI and RuleFit.     

On the Consistency of the Normalized Eight-Point Algorithm

A recently proposed argument to explain the improved performance of the eight-point algorithm that results from using normalized data (Chojnacki, W., et al. in IEEE Trans. Pattern Anal. Mach. Intell. 25(9):1172–1177, 2003) relies upon adoption of a certain model for statistical data distribution. Under this model, the cost function that underlies the algorithm operating on the normalized data is statistically more advantageous than the cost function that underpins the algorithm using unnormalized data. Here we extend this explanation by introducing a more refined, structured model for data distribution. Under the extended model, the normalized eight-point algorithm turns out to be approximately consistent in a statistical sense. The proposed extension provides a link between the existing statistical rationalization of the normalized eight-point algorithm and the approach of Mühlich and Mester for enhancing total least squares estimation methods via equilibration. The paper forms part of a wider effort to rationalize and interrelate foundational methods in vision parameter estimation.     

Finite rotations in the description of continuum deformation

Finite rotations in continuum mechanics are described by means of either a proper orthogonal tensor or finite rotation vectors. Some algebraic relations concerning the finite rotations are reviewed. Formulae expressing them in terms of displacements are given. Along each of the curvilinear coordinate lines the finite rotations are shown to satisfy some systems of the linear first-order differential equations. Each system of the equations is presented in four different but equivalent forms associated with an intermediate stretched basis or with an intermediate rotated basis. Integrability conditions of the system of equations provide various alternative forms of compatibility conditions in continuum mechanics. The displacement field is expressed through the stretch and rotation fields in the form of three successive line integrals. The formula describes the displacements to within a constant finite translation and a constant finite rotation. The procedure proposed here generalizes the formula derived by Cesàro (1906) within the classical linear theory of elasticity.     

Impossible and ambiguous shading patterns

A smooth object depicted in a monochrome image will often exhibit brightness variation, or shading. A problem much studied in computer vision has been that of how object shape may be recovered from image shading. When the imaging conditions are such that an overhead point-source illuminates a smooth Lambertian surface, the problem may be formulated as that of finding a solution to an eikonal equation. This article will focus on the existence and uniqueness of such solutions, reporting recent results obtained. With regard to existence, shading patterns are exhibited for which there is no corresponding object shape. Specifically, a necessary and sufficient condition is presented for a circularly symmetric eikonal equation to admit exclusively unbounded solutions; additionally, a sufficient condition is given for an eikonal equation to have no solution whatsoever. In connection with uniqueness, we consider eikonal equations, defined over a disc, such that the Euclidean norm of the gradient of any solution is circularly symmetric, vanishes exactly at the disc center, and diverges to infinity as the circumference of the disc is approached. Contrary to earlier influential work, a class of such eikonal equations is shown to possess simultaneously circularly symmetric and noncircularly symmetric bounded smooth solutions.     

Hydrophilic polycarbonate chips for generation of oil-in-water (O/W) and water-in-oil-in-water (W/O/W) emulsions

Modification of the surfaces of polycarbonate (PC) with the use of a solution of tin (II) chloride renders them hydrophilic. The surface draping is stable against exposure to water and to alcohols. Exposure to alkanes reduces but does not diminish the effect. The method is compatible—in using the same solvent and temperature—with the hydrophobic modification of PC (Jankowski et al. in Lab Chip 11:1151–1156, 2011). The combination of these methods makes it possible to generate single and multiple monodisperse emulsions with the use of flow-focusing junctions in systems made in PC—a material that is suitable for fabrication of multilayer, high-throughput microfluidic devices.     

An analysis of a contention resolution algorithm

Summary A single multiaccess channel is studied with the outcome of a transmission being either idle, success, or collision (ternary channel). Packets involved in a collision must be retransmitted, and an efficient way to solve a collision is known in the literature as Gallager-Tsybakov-Mikhailov algorithm. Performance analysis of the algorithm is quite hard. In fact, it bases on a numerical solution of some recurrence equations and on a numerical evaluation of some series. The obvious drawback of it is lack of insight into the behavior of the algorithm. We shall present a new approach of looking at the algorithm and discuss some attempts of analyzing its performance. In particular, expected lengths of a resolution interval and a conflict resolution interval as well as throughput of the algorithm will be discussed using asymptotic approximation and a small input rate approximation.     

Surface grain structure evolution in hot rolling of 6061 aluminum alloy

During hot deformation of Al–Mg–Si alloys, a non-uniform microstructure is developed due to differences in localized strain. Physical simulations were performed to examine the effects of processing parameters such as deformation temperature, total reduction, rolling schedule and alloying additions on the grain structure evolution. Overall, the kinetics of recrystallization followed the traditional trends predicted by JMAK kinetics. Electron backscatter diffraction (EBSD) was used to quantify the difference in recrystallization kinetics at the sample surface and mid-thickness. The results showed that the surface showed elevated kinetics when compared to the sample mid-thickness.     

Coupling monolayers for protection of microelectronic circuits

Two polymeric coatings, a silicone gel (Dow Corning 6646) and an epoxy resin (Dexter FP 4402), were glob-top coated onto representative microelectronic circuits, AT&T Triple Track Testers (TTTs), and subjected to the Pressure Cooker Test (PCT). Coupling monolayers were self- assembled on the TTTs prior to encapsulation to improve the moisture protection capabilities of the coatings. Leakage current measurements were made to evaluate the effect of applied monolayers on the moisture protection capability. The moisture protection capability was assessed in short-term and long-term leakage current measurements. MHDA (16-mercaptohexadecanoic acid) and APS (gamma-aminopropyltriethoxysilane) monolayers, in combination with silicone gel and epoxy resin respectively, exhibited very good moisture protection performance.     

The role of ammonium citrate washing on the characteristics of mechanochemical–hydrothermal derived magnesium-containing apatites

The role of citrate washing on the physical and chemical characteristics of magnesium-substituted apatites (HAMgs) was performed. HAMgs were synthesized by a mechanochemical–hydrothermal route at room temperature in as little as 1 h, which is five times faster than our previous work. Magnesium-substituted apatites had concentrations as high as 17.6 wt% Mg with a corresponding specific surface area (SSA) of 216 m²/g. A systematic study was performed to examine the influence of increasing magnesium content on the physical and chemical characteristics of the reaction products. As the magnesium content increased from 0 to 17.6 wt%, magnesium-doped apatite crystallite size decreased from 12 to 8.8 nm. The Mg/(Mg + Ca) ratio in the product was enriched relative to that used for the reacting precursor solution. During mechanochemical–hydrothermal reaction, magnesium doped apatites co-crystallize with magnesium hydroxide. Citrate washing serves to remove the magnesium hydroxide phase. The concomitant increase in surface area results because of the removal of this phase. Possible mechanisms for magnesium hydroxide leaching are discussed to explain the measured trends.