A5082524584- Advanced Numerical Methods in Computational Mathematics
- Elasticity and Material Modeling
- Advanced Numerical Analysis Techniques
- Numerical methods in engineering
- Contact Mechanics and Variational Inequalities
- Advanced Mathematical Modeling in Engineering
- Advanced Materials and Mechanics
- Distributed and Parallel Computing Systems
- Computational Geometry and Mesh Generation
- Computer Graphics and Visualization Techniques
- Computational Fluid Dynamics and Aerodynamics
- Orthopaedic implants and arthroplasty
- Matrix Theory and Algorithms
- Machine Learning in Bioinformatics
- Protein Structure and Dynamics
- Cellular Mechanics and Interactions
- Structural Analysis and Optimization
- Numerical methods in inverse problems
- Parallel Computing and Optimization Techniques
- Composite Material Mechanics
- Electromagnetic Simulation and Numerical Methods
- Numerical methods for differential equations
- Soil and Unsaturated Flow
- Advanced Optimization Algorithms Research
- RNA and protein synthesis mechanisms
TU Dresden
2015-2024
Freie Universität Berlin
2005-2019
Institute of Numerical Mathematics
2019
University of Vienna
2019
ETH Zurich
2019
RWTH Aachen University
2013-2016
Numerical Method (China)
2016
Karlsruhe Institute of Technology
2014-2015
Gesellschaft Fur Mathematik Und Datenverarbeitung
2014
Berlin Heart (Germany)
2010-2012
Summary: ROCR is a package for evaluating and visualizing the performance of scoring classifiers in statistical language R. It features over 25 measures that can be freely combined to create two-dimensional curves. Standard methods investigating trade-offs between specific are available within uniform framework, including receiver operating characteristic (ROC) graphs, precision/recall plots, lift charts cost integrates tightly with R's powerful graphics capabilities, thus allowing highly...
Abstract Motivation: In life sciences, interpretability of machine learning models is as important their prediction accuracy. Linear are probably the most frequently used methods for assessing feature relevance, despite relative inflexibility. However, in past years effective estimators relevance have been derived highly complex or non-parametric such support vector machines and RandomForest (RF) models. Recently, it has observed that RF biased a way categorical variables with large number...
We compared several statistical learning methods for the prediction of HIV coreceptor use from clonal third hypervariable (V3) loop sequences, and evaluated improved their effectiveness on clinical samples.Support vector machines (SVM), artificial neural networks, position-specific scoring matrices (PSSM) mixtures localized rules were estimated tested using 10x ten-fold cross-validation a dataset consisting 1,100 matched genotype-phenotype pairs 332 patients. Different SVMs also trained...
This paper presents the basic concepts and module structure of Distributed Unified Numerics Environment reflects on recent developments general changes that happened since release first Dune version in 2007 main papers describing state Bastian etal. (2008a, 2008b). discussion is accompanied with a description various advanced features, such as coupling domains cut cells, grid modifications adaptation moving domains, high order discretizations node level performance, non-smooth multigrid...
Nanophase mixtures, leveraging the complementary strengths of each component, are vital for composites to overcome limitations posed by single elemental materials. Among these, metal-elastomer nanophases particularly important, holding various practical applications stretchable electronics. However, methodology and understanding nanophase mixing metals elastomers limited due difficulties in blending caused thermodynamic incompatibility. Here, we present a controlled method using kinetics mix...
The Dune project has released version 2.4 on September 25, 2015. This paper describes the most significant improvements, interface and other changes for core modules Dune- Common, Dune-Geometry, Dune-Grid, Dune-ISTL, Dune-LocalFunctions.
HIV-1 cell entry commonly uses, in addition to CD4, one of the chemokine receptors CCR5 or CXCR4 as coreceptor. Knowledge coreceptor usage is critical for monitoring disease progression well supporting therapy with novel drug class antagonists. Predictive methods inferring based on third hypervariable (V3) loop region viral gene coding envelope protein gp120 can provide us these facilities while avoiding expensive phenotypic tests. All simple heuristics (such 11/25 rule) statistical learning...
We derive the quasiconvex relaxation of Biot-type energy density $\lVert\sqrt{\operatorname{D}\varphi^T \operatorname{D}\varphi}-I_2\rVert^2$ for planar mappings $\varphi\colon\mathbb{R}^2\to \mathbb{R}^2$ in two different scenarios. First, we consider case $\operatorname{D}\varphi\in\textrm{GL}^+(2)$, which can be expressed as squared Euclidean distance $\operatorname{dist}^2(\operatorname{D}\varphi,\textrm{SO}(2))$ to special orthogonal group $\textrm{SO}(2)$. then allow with arbitrary...
Abstract Background In recent years protein structure prediction methods using local information have shown promising improvements. The quality of new fold predictions has risen significantly and in recognition incorporation led to improvements the accuracy results. We developed a method be integrated into either or methods. For each sequence window predicts probability estimates for attain particular structures from set predefined candidates. first step is define representatives based on...
Abstract Motivation: An approach for identifying similarities of protein–protein binding sites is presented. The geometric shape a site described by computing feature vector based on moment invariants. In order to search similarities, vectors are compared. Similar indicate with similar shapes. Results: validated representative set sites, extracted from the SCOPPI database. When querying set, we known among 2819 sites. A median area under ROC curve 0.98 observed. For half queries, identified...
Abstract We introduce geodesic finite elements as a new way to discretize the non‐linear configuration space of geometrically exact Cosserat rod. These naturally generalize standard one‐dimensional spaces functions with values in Riemannian manifold. For special orthogonal group, our approach reproduces interpolation formulas Crisfield and Jelenić. Geodesic are conforming lead objective path‐independent problem formulations. for general manifolds, discuss relationship between coefficient...
SUMMARY We introduce geodesic finite elements as a conforming way to discretize partial differential equations for functions v : Ω → M , where is an open subset of and Riemannian manifold. These naturally generalize standard first‐order Euclidean spaces. They also the proposed d = 1 in previous publication author. Our formulation equivariant under isometries and, hence, preserves objectivity continuous problem formulations. concentrate on that can be formulated minimization problems....
Abstract The next generation of sensors requires a simple yet compact lab on chip‐based precise optical detection mechanism where data interpretation can be achieved with minimum effort. Hereby, cost‐efficient strategies manufacturing both propagating surface plasmon polariton (SPP) and localized resonance (LSPR) flexible platforms are explored via mechanical instabilities oblique‐angled metal evaporation. Centimeter scaled dielectric grating structures produced by plasma oxidation...
We assess the variability of protein function in sequence and structure space. Various regions this space exhibit considerable difference local conservation molecular function. analyze capture by means logistic curves. Based on analysis, we propose a method for predicting query with known but unknown The prediction is rigorously assessed compared previously published predictor. Furthermore, apply to 500 functionally unannotated PDB structures discuss selected examples. proposed approach...
We derive and analyze a solver-friendly finite element discretization of time discrete Richards equation based on Kirchhoff transformation. It can be interpreted as classical in physical variables with nonstandard quadrature points. Our approach allows for nonlinear outflow or seepage boundary conditions Signorini type. show convergence the saturation and, nondegenerate case, pressure. The associated algebraic problems formulated convex minimization therefore, solved efficiently by monotone...
We generalize geodesic finite elements to obtain spaces of higher approximation order. Our approach uses a Riemannian centre mass with signed measure. prove well-definedness this new under suitable conditions. As side product, we can define for non-simplex reference such as cubes and prisms. smoothness the interpolation functions various invariance properties. Numerical tests show that optimal convergence orders discretization error known from linear theory are obtained also in nonlinear setting.