Search

AbstractThis article proposes an adaptively bounded gradient descent (ABGD) algorithm for group elastic net penalized regression. Unlike previously proposed algorithms, the proposed algorithm adaptively bounds the Fisher information matrix, which results in a flexible and stable computational framework. In particular, the proposed algorithm (i) does not require orthogonalization of the predictors, and (ii) can be easily applied to any combination of exponential family response distribution and link function. The proposed algorithm is implemented in the grpnet R package (available from CRAN), which implements the approach for common response distributions (Gaussian, binomial, and Poisson), as well as several response distributions not previously considered in the group penalization literature (i.e., multinomial, negative binomial, gamma, and inverse Gaussian). Simulated and real data examples demonstrate that the proposed algorithm is as or more efficient than existing methods for Gaussian, binomial, and Poisson distributions. Furthermore, using two genomic examples, I demonstrate how the proposed algorithm can be applied to high-dimensional multinomial regression problems with grouped predictors. R code to reproduce the results is included as supplementary materials. [ABSTRACT FROM AUTHOR]

Preoperative compression of middle cerebellar peduncle (MCP) is often observed in vestibular schwannomas. Its re-expansion is expected after tumour resection, however, frequently its thickness remains unchanged or undergoes further atrophy. Similarly, increased MCP FLAIR signal is often observed and thought to be associated with intraoperative MCP injury. This study investigates the dynamics of MCP FLAIR signal changes over time and their implications in long-term MCP atrophy. Retrospective analysis of patients operated between 2011 and 2019 was performed. Measurements of FLAIR signals and MCP thickness were performed preoperatively, postoperatively and at follow-up. 28 patients (15 females, mean age 51.94 years) were included. The mean follow-up was 23.98 months. The mean tumour size was 2.99 cm. The MCP FLAIR signal was elevated preoperatively in 10 (35.7%) patients and further increased postoperatively in 22 (78.6%), followed by its decrease at follow up (7 patients, 25%). An immediate postoperative re-expansion of middle cerebellar peduncle was observed in 24 (85.7%) patients. No association between tumour size and preoperative FLAIR was established, however tumour size was negatively associated with the MCP thickness. A significant negative association between a postoperative FLAIR and follow-up thickness (p < 0.001) was noted, even if controlling for tumour size and both tumour size and preoperative MCP thickness. In patients with vestibular schwannomas undergoing surgical resection, the middle cerebellar peduncle FLAIR signal seems to associated with long term thickness of MCP, regardless of its initial size, however does not seem to correlate with the clinical outcome. [ABSTRACT FROM AUTHOR]

We studied the electrical conductivity of a bilayer specimen of N,N′-bis(naphthalen-1-yl)-N,N′-bis(phenyl)benzidine (α-NPD) and 1,3-bis(N-carbazolyl) benzene (mCP). The conductivity of indium tin oxide ITO/(heptadecafluoro-1,1,2,2-tetrahydrodecyl)triethoxysilane (FSAM)/α-NPD(50 nm)/Al can be explained by the space-charge-limited current (SCLC) model. In an α-NPD(50 nm)/mCP(10 nm) specimen, the current is similar to that in an α-NPD(50 nm) specimen despite the existence of a hole barrier between the α-NPD and mCP. The current in an α-NPD/mCP specimen is also proportional to the square of the voltage. We propose a conduction model that combines SCLC in the α-NPD layer and tunneling in the mCP layer. [ABSTRACT FROM AUTHOR]

In this paper, a family of coordinate majorization descent algorithms are proposed for solving the nonconvex penalized learning problems including SCAD and MCP estimation. In the coordinate majorization descent algorithms, each coordinate descent step is replaced with a coordinate-wise majorization descent operation, and the convergence of the algorithms are discussed in linear models. In addition, we apply the algorithms to the Logisitic models. Our simulation study and data examples indicate that the coordinate majorization descent algorithms can select the real model with a higher probability and the model is sparse, also the algorithms improve the accuracy of the parameter estimation with SCAD and MCP penalties. [ABSTRACT FROM AUTHOR]

Quantifying joint deformity in people with rheumatoid (RA) and psoriatic arthritis (PsA) remains challenging. Here, we demonstrate a new method to measure bone erosions and abnormal periosteal growths, based on the difference between a predicted healthy and actual diseased joint surface. We optimized the method by creating and measuring artificial bone erosions and growths. Then we measured 46 healthy and diseased patient surfaces. We found average sensitivity errors of ≤0.27 mm when measuring artificial erosions and growths. Patients had significantly more bone erosion than healthy subjects. Surface based outcomes are a novel way to interpret and quantify bone changes in PsA and RA. [ABSTRACT FROM AUTHOR]

Friedman et al. proposed the fused lasso signal approximator (FLSA) to denoise piecewise constant signals by penalizing the ℓ1 differences between adjacent signal points. In this article, we propose a new method, referred to as the fused-MCP, by combining the minimax concave penalty (MCP) with the fusion penalty. The fused-MCP performs better than the FLSA in maintaining the profile of the original signal and preserving the edge structure. We show that, with a high probability, the fused-MCP selects the right change-points and has the oracle property, unlike the FLSA. We further show that the fused-MCP achieves the same l2 error rate as the FLSA. We develop algorithms to solve fused-MCP problems, either by transforming them into MCP regression problems or by using an adjusted majorization-minimization algorithm. Simulation and experimental results show the effectiveness of our method. Supplementary material for this article is available online. [ABSTRACT FROM AUTHOR]

Territorial, socially monogamous species actively defend their home range against conspecifics to maintain exclusive access to resources such as food or mates. Primates use scent marks and loud calls to signal territory occupancy and limit the risk of intergroup encounters, maximizing their energetic balance. Indri indri is a little-studied territorial, socially monogamous singing primate living in family groups. The groups announce territory occupancy with long-distance calls, and actively defend their territories from conspecific intruders. This work includes data collected in three forests in Madagascar on 16 indri groups over up to 5 years. We aimed (1) to estimate the extent of territories using minimum convex polygon (MCP), implementing minimum sampling effort requirements; (2) to quantify territorial exclusivity, measuring the overlap between territories; and (3) to evaluate the intergroup encounter rate and to quantify the dynamics of group encounters. Our results showed that indris range evenly within exclusive small territories with no or little overlap. Intergroup encounters are rare (0.05 encounters per day), and are located on the periphery of the territories. Disputes were mostly solved with vocal confrontation and only in 13% of the cases ended in physical fights. This frame underlines a cost–benefit explanation of territoriality, favouring a strategy that efficiently limits overlap and avoids costly intergroup encounters. We hypothesize that territorial behaviour in indri is related to mate-guarding strategy and that vocal behaviour plays a fundamental role in regulating intergroup spacing dynamics. [ABSTRACT FROM AUTHOR]

We propose and study a new iterative coordinate descent algorithm (QICD) for solving nonconvex penalized quantile regression in high dimension. By permitting different subsets of covariates to be relevant for modeling the response variable at different quantiles, nonconvex penalized quantile regression provides a flexible approach for modeling high-dimensional data with heterogeneity. Although its theory has been investigated recently, its computation remains highly challenging whenpis large due to the nonsmoothness of the quantile loss function and the nonconvexity of the penalty function. Existing coordinate descent algorithms for penalized least-squares regression cannot be directly applied. We establish the convergence property of the proposed algorithm under some regularity conditions for a general class of nonconvex penalty functions including popular choices such as SCAD (smoothly clipped absolute deviation) and MCP (minimax concave penalty). Our Monte Carlo study confirms that QICD substantially improves the computational speed in thep≫nsetting. We illustrate the application by analyzing a microarray dataset. [ABSTRACT FROM PUBLISHER]

In this article, we develop a generalized penalized linear unbiased selection (GPLUS) algorithm. The GPLUS is designed to compute the paths of penalized logistic regression based on the smoothly clipped absolute deviation (SCAD) and the minimax concave penalties (MCP). The main idea of the GPLUS is to compute possibly multiple local minimizers at individual penalty levels by continuously tracing the minimizers at different penalty levels. We demonstrate the feasibility of the proposed algorithm in logistic and linear regression. The simulation results favor the SCAD and MCP’s selection accuracy encompassing a suitable range of penalty levels. [ABSTRACT FROM AUTHOR]

New red light-emitting phosphorescent devices using the mixed host of N,N′-dicarbazolyl-3,5-benzene [mCP] and bis(10-hydroxy-benzo[h]quinolinato)beryllium [Bebq2] doped with 5 vol.% RP411(red phosphorescent dye from SFC Co.) in the emissive layer were fabricated and evaluated according to the mixture ratio of x in a host of mCP(1−x) Bebq2(x). Among the fabricated devices, the maximum quantum efficiency was obtained in the device with in a host of x = 0.75 and the highest luminance at a given voltage was obtained in the device with a in a host of x = 0.25. On the other hand, the product of luminance and quantum efficiency showed the best characteristics in the device with a host of mCP(0.5) Bebq2(0.5). The peak wavelength in the electroluminescence spectra and color coordinates on the Commission Internationale de I’Eclairage(CIE) chart of the fabricated devices with x ≥ 0.25 in a host of mCP(1−x) Bebq2(x) were 627 nm and (0.66, 0.34), respectively. [ABSTRACT FROM PUBLISHER]

The objectives of this article are to illustrate theoretically and empirically how to incorporate an ad valorem tariff in spatial equilibrium models and show the equivalence of three approaches (primal, dual and MCP) of spatial equilibrium trade modelling. The article lays out in detail the theory, empirical applications and the results for all three approaches so that trade modellers can easily follow and apply them in their work. The primal approach is commonly used by spatial equilibrium modellers, though the dual approach corresponds to the textbook depiction of consumer surplus, producer surplus and gains from trade. The Mixed Complementarity Problem (MCP) is not commonly used by spatial equilibrium modellers, though it is relatively simpler if the modeller knows the equilibrium conditions. This article also presents advantages and disadvantages of each approach. [ABSTRACT FROM PUBLISHER]

Political parties are regarded as key institutions in the consolidation of democracy. The extent to which a political party is institutionalized is said to be a critical aspect that determines its effectiveness and the contribution it makes toward democracy. Despite this profound discovery, only a few studies have focused on assessing the institutionalization of individual political parties as opposed to those on party system institutionalization. In attempting to bridge the gap, the paper investigates the extent to which political parties in Malawi are institutionalized. The study, taking advantage of different party institutionalization models, focused on analyzing and comparing Malawi political parties in the name of United Democratic Front (UDF) and Malawi Congress Party (MCP) on dimensions of systemness, autonomy, value infusion and reification based on the Randall and Svasand (2002) party institutionalization model. The study utilized data collected through field interviews with officials at national, regional and district levels as well as data from secondary sources. The methodology entailed combined uses of qualitative and quantitative tools. The study findings indicate that political parties in Malawi are poorly institutionalized, which significantly impacts on their functioning. Whereas UDF scores slightly better than MCP on decisional autonomy, the party lags behind in systemness, value infusion and reification. Although the study attempts to portray a differentiated picture, on the whole, the results imply the parties' limited state of institutionalization hence constrains their contribution to democratic consolidation. [ABSTRACT FROM AUTHOR]

Novel routing paradigms, based on policies and quality of service (QoS) requirements, have been proposed for the Internet over the last decade. Constraint-based path selection algorithm is of utmost importance in the field of QoS routing. An Internet service provider wants to find paths for QoS flows that optimize the network resources and satisfy the user requirements. But, unfortunately the problem of finding a path with multiple constraints such as delay, jitter, bandwidth, and packet loss is a Nondeterministic polynomial (NP) complete problem. Hence, various heuristics have been proposed to find a path satisfies that more than one constraint. All those heuristics find feasible path using link-state information, which are considered to be static throughout. These heuristics do not consider the dynamic parameter such as the queuing situation at a given node. Hence, in this paper, a heuristic called dynamic multi constraint multi path routing algorithm (DMCMPRA), which incorporates a better cost model using dynamic variables (viz., buffer occupancy at a node to estimate the link cost) to compute the best path for the QoS How and to balance the load among alternate feasible paths to the destination, is proposed. DMCMPRA uses the self-adaptive multi-constrained routing algorithm (SAMCRA) to find the k QoS aware paths. Then the link-disjoint multiple constraints routing algorithm (DIMCRA) is used to find all the in link disjoint paths among these k paths. The proposed new cost model is used to select the best path from the n link disjoint paths. On congestion notification, load balancing is achieved by selecting an alternate feasible path from then n - 1 link disjoint path using the same cost model Computer simulations show that the performance of DMCMPRA is better than SAMCRA in terms of the metrics such as percentage of successfully received packets, average delay, etc. [ABSTRACT FROM AUTHOR]

A laboratory incubation experiment using soil columns was conducted to study the effects of monocalcium phosphate (MCP) and potassium chloride (KCl) on soil pH changes in fertilizer microsites with two Chinese soils. Mixtures of two fertilizers at two rates (0 and 0.26 g per column) were added to the surface of soil cylinders. The results shown that both fertilizers significantly decreased soil pH after 7 d and 28 d, which declined with time and distance from fertilizer site. Compared with KCl alone, the soil pH decrease close to the fertilizer site induced by applied KCl was slowed down in the acidic red soil by MCP addition but was promoted in the calcareous soil. Compared with MCP alone, the application of KCl with MCP had greater effects of reducing pH in both soils. The magnitude and extent of soil pH changes were mostly contributed by KCl in the KCl plus MCP treatment. [ABSTRACT FROM AUTHOR]