

 Name (Last, First), mail  FZJ Institute/Group  Your ML expertise  Your type of data  Your needs  expire date (default 25.6.2019) 



::::::



 Jitsev, Jenia j.jitsev@fzjuelich.de  JSC Deep Learning Research Lab (CSTDL)  Unsupervised Learning, Reinforcement Learning, Recurrent Hierarchical WinnerTakeAll Networks, Generative Models, Biological Neural Networks, Openend learning / Learning to learn  Images, Time Series, Virtual Environments (e.g, OpenGym)  Large scale scientific data sets (material science, high throughput genomics/proteomics, biotechnology, etc)  none 



 Wenzel, Susanne s.wenzel@fzjuelich.de  INM1 Big Data Analytics Group (BDA)  Markov Marked Point Processes, rjMCMC    14.02.2020 



Dickscheid, Timo t.dickscheid@fzjuelich.de  INM1 Big Data Analytics Group (BDA)  Deep Learning with ConvNets, Image Segmentation, Image Feature detection, Markov Random Fields, Clustering, Analogies to the human brain Microscopic resolution 2D and 3D images    none 



 Kraus, Jiri jkraus@nvidia.com  NVIDIA    Want to learn the ML/DL needs of scientist at FZJ  none 



Zimmermann, Olav olav.zimmermann@fzjuelich.de  JSC Simlab Biology  Supervised learning (SVM and its variants), Dimension Reduction (Isomap et al, clustering), Metaheuristics, basics of sequence learning methods (CRF, LSTM), bioinformatics  biological sequence data, molecular structure data, experimental data (2dim to ndim or graphs), unstructured data  ML for hypertoroidal output spaces in depth understanding of seq2seq methods, representation of world knowledge for learning methods that work with nondifferentiable loss  



Herten, Andreas a.herten@fzjuelich.de  JSC NVIDIA Application Lab  Software installation   Overview/support of ML/DL software in Jülich  



 Göbbert, Jens Henrik j.goebbert@fzjuelich.de  JSC CrossSectionalTeam Visualization  Basic DL with Keras  turbulent flows (3D)  Collaborations/support for enabling DL with on HPC with Jupyter  none 



Hermanns, MarcAndré m.a.hermanns@fzjuelich.de  JSC CrossSectionalTeam Parallel Performance  none yet (testing with Keras atm.)  Performance Data (profile and trace)  Identify or estimate performance phenomena  none 



 Wagner, Christian c.wagner@fzjuelich.de  PGI3 Molecular manipulation lab  general overview over ML principles, PCA, background on methods to combine ML and comptational chemistry, reinforcement learning with NNs  (hysteretic) scalars along (x,y,z), trajectories, computational, chemistry data, scalar time series  easy access to ML expertise, possibility to discuss (and potentially solve) individual ML problems / tasks at detailed level, potentially in the frame of a collaboration  none 



 Name (Last, First), mail  FZJ Institute/Group  Your ML expertise  Your type of data  Your needs 



:::::



 Jitsev, Jenia j.jitsev@fzjuelich.de  JSC Deep Learning Research Lab (CSTDL)  Unsupervised Learning, Reinforcement Learning, Recurrent Hierarchical WinnerTakeAll Networks, Generative Models, Biological Neural Networks, Openend learning / Learning to learn  Images, Time Series, Virtual Environments (e.g, OpenGym)  Large scale scientific data sets (material science, high throughput genomics/proteomics, biotechnology, etc) 



 Wenzel, Susanne s.wenzel@fzjuelich.de  INM1 Big Data Analytics Group (BDA)  Markov Marked Point Processes, rjMCMC   



Dickscheid, Timo t.dickscheid@fzjuelich.de  INM1 Big Data Analytics Group (BDA)  Deep Learning with ConvNets, Image Segmentation, Image Feature detection, Markov Random Fields, Clustering, Analogies to the human brain Microscopic resolution 2D and 3D images   



 Kraus, Jiri jkraus@nvidia.com  NVIDIA    Want to learn the ML/DL needs of scientist at FZJ 



Zimmermann, Olav olav.zimmermann@fzjuelich.de  JSC Simlab Biology  Supervised learning (SVM and its variants), Dimension Reduction (Isomap et al, clustering), Metaheuristics, basics of sequence learning methods (CRF, LSTM), bioinformatics  biological sequence data, molecular structure data, experimental data (2dim to ndim or graphs), unstructured data  ML for hypertoroidal output spaces in depth understanding of seq2seq methods, representation of world knowledge for learning methods that work with nondifferentiable loss 



Herten, Andreas a.herten@fzjuelich.de  JSC NVIDIA Application Lab  Software installation   Overview/support of ML/DL software in Jülich 



 Göbbert, Jens Henrik j.goebbert@fzjuelich.de  JSC CrossSectionalTeam Visualization  Basic DL with Keras  turbulent flows (3D)  Collaborations/support for enabling DL with on HPC with Jupyter 



Hermanns, MarcAndré m.a.hermanns@fzjuelich.de  JSC CrossSectionalTeam Parallel Performance  none yet (testing with Keras atm.)  Performance Data (profile and trace)  Identify or estimate performance phenomena 



 Wagner, Christian c.wagner@fzjuelich.de  PGI3 Molecular manipulation lab  general overview over ML principles, PCA, background on methods to combine ML and comptational chemistry, reinforcement learning with NNs  (hysteretic) scalars along (x,y,z), trajectories, computational, chemistry data, scalar time series  easy access to ML expertise, possibility to discuss (and potentially solve) individual ML problems / tasks at detailed level, potentially in the frame of a collaboration 



 Hagemeier, Björn b.hagemeier@fzjuelich.de  JSC Project Helmholtz Analytics Framework  Basic ML methods    



 Stadtler, Scarlet s.stadtler@fzjuelich.de  JSC Federated Systems and Data Division  Absolute Beginner (At JSC courses in ML and DL)  Meteorological four dimensional Data (space and time)  meet ML experts, possibility to discuss individual DL problems  



 Goergen, Klaus k.goergen@fzjuelich.de  IBG3 Integrated Modelling  various supervised and unsupervised methods, PCA, CCA, SOMs; not used recently though  Regional climate model outputs, meteorological observations  big data (200300TB) capable data analytics frameworks  none 



 Krajsek, Kai k.krajsek@fzjuelich.de  JSC HPC in Neuroscience  Supervised Learning Unsupervised Learning MetaLearning, Inverse Modelling with ML, Deep Learning in Computer Vision, Probabilistic Inference, Gaussian Processes  Meteorological model output (4D spacetime volumes), Diffusion MRI data  



 Pleiter, Dirk d.pleiter@fzjuelich.de  JSC Technology Department  Architectures optimized for DL, requirements analysis    none 



 Arasan, Durai d.arasan@fzjuelich.de  INM1 Connectivity  Support Vector Machines, Random Forests, Neural Networks  3D MRI data  Developing ML algorithms for neuroimaging data, Learning Deep Learning, feature engineering and optimization  none 



 Stadtler, Scarlet s.stadtler@fzjuelich.de  JSC Federated Systems and Data Division  Absolute Beginner (At JSC courses in ML and DL)  Meteorological four dimensional Data (space and time)  meet ML experts, possibility to discuss individual DL problems 



 Goergen, Klaus k.goergen@fzjuelich.de  IBG3 Integrated Modelling  various supervised and unsupervised methods, PCA, CCA, SOMs; not used recently though  Regional climate model outputs, meteorological observations  big data (200300TB) capable data analytics frameworks 



 Krajsek, Kai k.krajsek@fzjuelich.de  JSC HPC in Neuroscience  Supervised Learning Unsupervised Learning MetaLearning, Inverse Modelling with ML, Deep Learning in Computer Vision, Probabilistic Inference, Gaussian Processes  Meteorological model output (4D spacetime volumes), Diffusion MRI data 



 Pleiter, Dirk d.pleiter@fzjuelich.de  JSC Technology Department  Architectures optimized for DL, requirements analysis   



 Arasan, Durai d.arasan@fzjuelich.de  INM1 Connectivity  Support Vector Machines, Random Forests, Neural Networks  3D MRI data  Developing ML algorithms for neuroimaging data, Learning Deep Learning, feature engineering and optimization 



 Huysegoms, Marcel m.huysegoms@fzjuelich.de  INM1 Big Data Analytics Group (BDA)  Deep Learning with ConvNets, Markov Random Fields, Clustering  Microscopicresolution 2D and 3D cyto images  



Schultz, Martin m.schultz@fzjuelich.de  JSCFSD  Deep Learning Timeseries and videosequence analysis (IntelliAQ project)  Observational time series, numerical weather model data (gridded fields), geodata (gridded fields and point features  discussion forum to exchange experiences with specific architectures and methods, and to share ideas how to improve result 



Schiffer, Christian c.schiffer@fzjuelich.de  INM1 Big Data Analytics Group (BDA)  Deep Learning with ConvNets



Image Segmentation, Distributed DL on HPC with TensorFlow, Unsupervised Domain Adaptation  Microscopic resolution 2D images 



Campos, Lucas l.campos@fzjuelich.de  INM1 Big Data Analytics Group (BDA)  None; JSC courses on Machine Learning/Deep Learning. Weekend fun with SVM and Neural networks way back when  3D MRI data   