A5066354297- Computational Drug Discovery Methods
- Microstructure and Mechanical Properties of Steels
- Machine Learning in Materials Science
- Metal Alloys Wear and Properties
- Welding Techniques and Residual Stresses
- Metallurgy and Material Forming
- Protein Structure and Dynamics
- Planetary Science and Exploration
- Context-Aware Activity Recognition Systems
- Analytical Chemistry and Chromatography
- Geology and Paleoclimatology Research
- Astro and Planetary Science
- Metabolomics and Mass Spectrometry Studies
- Hydrogen embrittlement and corrosion behaviors in metals
- Magnetic Properties and Applications
- Aluminum Alloy Microstructure Properties
- Machine Learning in Bioinformatics
- Microstructure and mechanical properties
- Industrial Vision Systems and Defect Detection
- Optical measurement and interference techniques
- Human Mobility and Location-Based Analysis
- IoT-based Smart Home Systems
- Cholinesterase and Neurodegenerative Diseases
- Online Learning and Analytics
- Non-Destructive Testing Techniques
MS-Schramberg (Germany)
2021-2023
TRUMPF (Germany)
2021-2023
University of Kassel
2015-2018
Freie Universität Berlin
2015-2017
Comtech Telecommunications (United States)
2015
Mannesmann (Germany)
2013-2014
Austrian Foundry Research Institute
2014
University of Tübingen
2008-2013
Fraunhofer Institute for Material and Beam Technology
2013
Bernstein Center for Computational Neuroscience Tübingen
2010-2013
Ligand-based virtual screening experiments are an important task in the early drug discovery stage. An ambitious aim each experiment is to disclose active structures based on new scaffolds. To perform these "scaffold-hoppings" for individual problems and targets, a plethora of different similarity methods diverse techniques were published last years. The optimal assignment approach molecular graphs, successful method field quantitative structure-activity relationships, has not been tested as...
The decomposition of a chemical graph is convenient approach to encode information the corresponding organic compound. While several commercial toolkits exist molecules as so-called fingerprints, only few open source implementations are available. aim this work introduce library for exactly defined molecular decompositions, with strong focus on application these features in machine learning and data mining. It provides options such search depth, distance cut-offs, atom- pharmacophore typing....
Abstract The temperature dependence of the martensite formation and mechanical properties three high alloyed Cr‐Mn‐Ni as‐cast steels with varying Ni contents were studied. results showed that M s d temperatures decrease increasing nickel contents. Therefore strain‐induced formation, TRIP effect anomaly elongations occurs at lower temperatures. steel 3% shows a stress induced dynamic strain aging. Depending on content TWIP additionally to in investigated steels. study was performed by using...
Model-based virtual screening plays an important role in the early drug discovery stage. The outcomes of high-throughput screenings are a valuable source for machine learning algorithms to infer such models. Besides strong performance, interpretability model is desired property guide optimization compound later stages. Linear support vector machines showed have convincing performance on large-scale data sets. goal this study present heat map molecule coloring technique interpret linear Based...
The success of genome-wide association studies (GWAS) in deciphering the genetic architecture complex diseases has fueled expectations whether individual risk can also be quantified based on architecture. So far, disease prediction top-validated single-nucleotide polymorphisms (SNPs) showed little predictive value. Here, we applied a support vector machine (SVM) to Parkinson (PD) and type 1 diabetes (T1D), show that apart from magnitude effect size variants, heritability plays an important...
The virtual screening of large compound databases is an important application structural-activity relationship models. Due to the high structural diversity these data sets, it impossible for machine learning based QSAR models, which rely on a specific training set, give reliable results all compounds. Thus, consider subset chemical space in model applicable. approaches this problem that have been published so far mostly use vectorial descriptor representations define domain applicability...
Abstract The martensite start temperature (M s ), the austenite re‐transformation (A ) and finish f of six high alloyed Cr‐Mn‐Ni steels with varying Ni Mn contents in wrought as‐cast state were studied. aim this investigation is development relationships between M , A T 0 temperatures chemical composition a new type steels. investigations show that decrease increasing nickel manganese contents. depends on amount martensite. Regression equations for transformation are given. experimental...
Abstract Modern steel developments often use additional deformation mechanisms like the induced martensitic transformation (TRIP‐effect) and mechanical twinning (TWIP‐effect) to enhance elongation strength. Three high‐alloyed cast CrMnNi‐steels with different austenite stabilities were examined. Dependent on stability, TRIP‐effect TWIP‐effect found. A low stability causes a distinctive formation of α'‐martensite therefore strong strain hardening. The increase rate leads an in yield strength...
The technology of hairpin welding, which is frequently used in the automotive industry, entails high-quality requirements welding process. It can be difficult to trace defect back affected weld if a non-functioning stator detected during final inspection. Often, visual assessment cooled seam does not provide any information about its strength. However, based on behavior especially spattering, conclusions made quality weld. In addition, spatter component have serious consequences. this paper,...
Abstract Stress‐Temperature‐Transformation (STT) and Deformation‐Temperature‐Transformation (DTT) diagrams are suitable to characterize the TRIP (transformation‐induced plasticity) TWIP (twinning‐induced effect in steels. The triggering stresses for deformation‐induced microstructure transformation processes, characteristic temperatures, yield stress strength of steel plotted STT diagram as functions temperature. elongation values austenite, strain‐induced twins martensite formations shown...
A new generation of high alloyed cast CrMnNi-TRIP-steels was developed exhibiting strength (UTS) as well uniform elongation for a maximum energy-absorption.The nature the is formation -martensite in most deformed areas metastable austenitic steel, so that necking delayed.Because low stacking fault energy, due to alloying concept, deformation mainly accompanied by development bands coarse grains and further -martensite.Within this study, stress-strain curves will be evaluated corresponding...
High alloyed metastable austenitic or austenitic-martensitic steels show a strain induced formation of martensite during mechanical loading.These kinds are well known as material for rolled products.Based on the System Fe-Cr-Mn-Ni new generation cast with TRIP effect will be discussed.The investigations how properties and fraction formed influenced by varying Ni contents.The in state quite similar to those state.This is valid tensile compression loading.Under certain conditions, an...
The goal of this study was to adapt a recently proposed linear large-scale support vector machine binary cheminformatics classification problems and assess its performance on various benchmarks using virtual screening measures. We extended the library LIBLINEAR with state-of-the-art high-throughput metrics train classifiers whole large unbalanced data sets. formulation has an excellent if applied high-dimensional sparse feature vectors. An additional advantage is average complexity in number...
Sensors embedded in smartphones are an essential component for activity recognition. Even though the accelerometer is most widely used sensor, highest recognition accuracies obtained when using data collected from multiple sensors. However, use of sensors has adverse impact on energy consumption power-limited devices such as smartphones. In this paper, we present a new method to improve accuracy physical activities by only accelerometer. We utilize low-pass filter split acceleration into...
Abstract Weltweit verstärken sich die Bemühungen bei der Suche nach kostengünstigen austenitischen Stahlgüten mit hohem Energieabsorptionsvermögen. Dabei zeichnen verschiedene Forschungsrichtungen ab, auf Stähle TRIP‐, TWIP‐ und SBIP‐Effekt 1) oder entsprechenden Kombinationen konzentrieren. Am Institut für Eisen‐ Stahltechnologie TU Bergakademie Freiberg werden nichtrostende austenitische austenitisch‐martensitische Leichtbaustähle Kaltumform‐ Energieabsorptionsvermögen entwickelt im...
Metastable austenitic steels show excellent mechanical properties, such as high strength combined with ductility and toughness due to martensitic transformation under loading (transformation induced plasticity effect). A good energy consumption, and, in the case of high-alloyed metastable steels, a corrosion resistance, increase potential these materials for diverse applications, also regard safety requirements. Up now, numerous wrought alloys were investigated concerning behaviour,...
Abstract Stress‐Temperature‐Transformation (STT) and Deformation‐Temperature‐Transformation (DTT) diagrams are well‐suited to characterize the TRIP (transformation‐induced plasticity) TWIP (twinning‐induced effect in steels. The triggering stresses for deformation‐induced microstructure transformation processes, characteristic temperatures, yield stress strength of steel plotted STT diagram as functions temperature. elongation values austenite, strain‐induced twins martensite formations...
Abstract The mechanical behavior and microstructure evolution during deformation of novel austenitic Cr–Mn–Ni as‐cast steels with varied Ni content were investigated at various temperatures using static tensile tests, optical microscopy, the magnetic scale for detection ferromagnetic phase fraction. To summarize all knowledge about deformation‐induced processes, STT DTT diagrams developed steels. illustrate different mechanisms depending on temperature tension load, quantify elongation...
Machine learning (ML) is a key technology in smart manufacturing as it provides insights into complex processes without requiring deep domain expertise. This work deals with algorithms to determine 3D reconstruction from single 2D grayscale image. The potential of can be used for quality control because the height values contain relevant information that not visible data. Instead scans, estimated depth maps based on input image advantage simple setup and short recording time. Determining...