publications | Felix Kemeth

2024

Data-driven cold starting of good reservoirs

Lyudmila Grigoryeva, Boumediene Hamzi, Felix P. Kemeth, and 4 more authors

Physica D: Nonlinear Phenomena, 2024

Abs Bib Code

Using short histories of observations from a dynamical system, a workflow for the post-training initialization of reservoir computing systems is described. This strategy is called cold-starting, and it is based on a map called the starting map, which is determined by an appropriately short history of observations that maps to a unique initial condition in the reservoir space. The time series generated by the reservoir system using that initial state can be used to run the system in autonomous mode in order to produce accurate forecasts of the time series under consideration immediately. By utilizing this map, the lengthy “washouts” that are necessary to initialize reservoir systems can be eliminated, enabling the generation of forecasts using any selection of appropriately short histories of the observations.
@article{grigoryeva24_cold_start, title = {Data-driven cold starting of good reservoirs}, journal = {Physica D: Nonlinear Phenomena}, volume = {469}, pages = {134325}, year = {2024}, issn = {0167-2789}, doi = {https://doi.org/10.1016/j.physd.2024.134325}, url = {https://www.sciencedirect.com/science/article/pii/S0167278924002768}, author = {Grigoryeva, Lyudmila and Hamzi, Boumediene and Kemeth, Felix P. and Kevrekidis, Yannis and Manjunath, G. and Ortega, Juan-Pablo and Steynberg, Matthys J.}, keywords = {Reservoir computing, Generalized synchronization, Starting map, Forecasting, Path continuation, Dynamical systems}, }

2023

Black and gray box learning of amplitude equations: Application to phase field systems

Felix P. Kemeth, Sergio Alonso, Blas Echebarria, and 3 more authors

Phys. Rev. E, Feb 2023

Abs arXiv Bib HTML Code

We present a data-driven approach to learning surrogate models for amplitude equations and illustrate its application to interfacial dynamics of phase field systems. In particular, we demonstrate learning effective partial differential equations describing the evolution of phase field interfaces from full phase field data. We illustrate this on a model phase field system, where analytical approximate equations for the dynamics of the phase field interface (a higher-order eikonal equation and its approximation, the Kardar-Parisi-Zhang equation) are known. For this system, we discuss data-driven approaches for the identification of equations that accurately describe the front interface dynamics. When the analytical approximate models mentioned above become inaccurate, as we move beyond the region of validity of the underlying assumptions, the data-driven equations outperform them. In these regimes, going beyond black box identification, we explore different approaches to learning data-driven corrections to the analytically approximate models, leading to effective gray box partial differential equations.
@article{PhysRevE.107.025305, title = {Black and gray box learning of amplitude equations: Application to phase field systems}, author = {Kemeth, Felix P. and Alonso, Sergio and Echebarria, Blas and Moldenhawer, Ted and Beta, Carsten and Kevrekidis, Ioannis G.}, journal = {Phys. Rev. E}, volume = {107}, issue = {2}, pages = {025305}, numpages = {8}, year = {2023}, month = feb, publisher = {American Physical Society}, doi = {10.1103/PhysRevE.107.025305}, url = {https://link.aps.org/doi/10.1103/PhysRevE.107.025305}, dimensions = {true}, }

2022

NatComm
Learning Emergent Partial Differential Equations in a Learned Emergent Space

Felix P. Kemeth, Tom Bertalan, Thomas Thiem, and 4 more authors

Nature Communications, Feb 2022

Abs arXiv Bib HTML Code

We propose an approach to learn effective evolution equations for large systems of interacting agents. This is demonstrated on two examples, a well-studied system of coupled normal form oscillators and a biologically motivated example of coupled Hodgkin-Huxley-like neurons. For such types of systems there is no obvious space coordinate in which to learn effective evolution laws in the form of partial differential equations. In our approach, we accomplish this by learning embedding coordinates from the time series data of the system using manifold learning as a first step. In these emergent coordinates, we then show how one can learn effective partial differential equations, using neural networks, that do not only reproduce the dynamics of the oscillator ensemble, but also capture the collective bifurcations when system parameters vary. The proposed approach thus integrates the automatic, data-driven extraction of emergent space coordinates parametrizing the agent dynamics, with machine-learning assisted identification of an emergent PDE description of the dynamics in this parametrization.
@article{kemeth22_learn_emerg_partial_differ_equat, author = {Kemeth, Felix P. and Bertalan, Tom and Thiem, Thomas and Dietrich, Felix and Moon, Sung Joon and Laing, Carlo R. and Kevrekidis, Ioannis G.}, title = {Learning Emergent Partial Differential Equations in a Learned Emergent Space}, journal = {Nature Communications}, volume = {13}, number = {1}, pages = {3318}, year = {2022}, doi = {10.1038/s41467-022-30628-6}, url = {https://doi.org/10.1038/s41467-022-30628-6}, date_added = {Fri Jul 8 08:17:08 2022}, dimensions = {true}, }

An explanation of how mutant and wild-type mitochondria might stably co-exist in inherited mitochondrial diseases

Axel Kowald, Felix P. Kemeth, and Tom B. L. Kirkwood

PNAS Nexus, Sep 2022

Bib HTML

@article{10.1093/pnasnexus/pgac192,
  author = {Kowald, Axel and Kemeth, Felix P. and Kirkwood, Tom B. L.},
  title = {An explanation of how mutant and wild-type mitochondria might stably co-exist in inherited mitochondrial diseases},
  journal = {PNAS Nexus},
  volume = {1},
  number = {4},
  pages = {pgac192},
  year = {2022},
  month = sep,
  issn = {2752-6542},
  doi = {10.1093/pnasnexus/pgac192},
  url = {https://doi.org/10.1093/pnasnexus/pgac192},
  eprint = {https://academic.oup.com/pnasnexus/article-pdf/1/4/pgac192/48849713/pgac192.pdf},
}

2021

Initializing LSTM internal states via manifold learning

Felix P. Kemeth, Tom Bertalan, Nikolaos Evangelou, and 3 more authors

Chaos: An Interdisciplinary Journal of Nonlinear Science, Sep 2021

Abs arXiv Bib HTML

We present an approach, based on learning an intrinsic data manifold, for the initialization of the internal state values of long short-term memory (LSTM) recurrent neural networks, ensuring consistency with the initial observed input data. Exploiting the generalized synchronization concept, we argue that the converged, “mature” internal states constitute a function on this learned manifold. The dimension of this manifold then dictates the length of observed input time series data required for consistent initialization. We illustrate our approach through a partially observed chemical model system, where initializing the internal LSTM states in this fashion yields visibly improved performance. Finally, we show that learning this data manifold enables the transformation of partially observed dynamics into fully observed ones, facilitating alternative identification paths for nonlinear dynamical systems.
@article{kemeth2021initializing, author = {Kemeth, Felix P. and Bertalan, Tom and Evangelou, Nikolaos and Cui, Tianqi and Malani, Saurabh and Kevrekidis, Ioannis G.}, title = {Initializing {LSTM} internal states via manifold learning}, journal = {Chaos: An Interdisciplinary Journal of Nonlinear Science}, volume = {31}, number = {9}, pages = {093111}, year = {2021}, doi = {10.1063/5.0055371}, url = {https://doi.org/10.1063/5.0055371}, eprint = {https://doi.org/10.1063/5.0055371}, dimensions = {true}, }

Global Heteroclinic Rebel Dynamics Among Large 2-Clusters in Permutation Equivariant Systems

Bernold Fiedler, Sindre W. Haugland, Felix P. Kemeth, and 1 more author

SIAM Journal on Applied Dynamical Systems, Sep 2021

arXiv Bib HTML

@article{fiedler20_dynamics,
  author = {Fiedler, Bernold and Haugland, Sindre W. and Kemeth, Felix P. and Krischer, Katharina},
  title = {Global Heteroclinic Rebel Dynamics Among Large 2-Clusters in
                    Permutation Equivariant Systems},
  journal = {SIAM Journal on Applied Dynamical Systems},
  volume = {20},
  number = {3},
  pages = {1277-1319},
  year = {2021},
  doi = {10.1137/20M1361493},
  url = {https://doi.org/10.1137/20M1361493},
  eprint = {https://doi.org/10.1137/20M1361493},
}

2-Cluster fixed-point analysis of mean-coupled Stuart–Landau oscillators in the center manifold

Felix P. Kemeth, Bernold Fiedler, Sindre W. Haugland, and 1 more author

Journal of Physics: Complexity, Feb 2021

arXiv Bib HTML

@article{Kemeth_2021,
  author = {Kemeth, Felix P. and Fiedler, Bernold and Haugland, Sindre W. and Krischer, Katharina},
  title = {2-Cluster fixed-point analysis of mean-coupled Stuart–Landau oscillators
                    in the center manifold},
  journal = {Journal of Physics: Complexity},
  volume = {2},
  number = {2},
  pages = {025005},
  year = {2021},
  month = feb,
  publisher = {IOP Publishing},
  doi = {10.1088/2632-072X/abd0da},
  url = {https://dx.doi.org/10.1088/2632-072X/abd0da},
}

Global and local reduced models for interacting, heterogeneous agents

Thomas N. Thiem, Felix P. Kemeth, Tom Bertalan, and 2 more authors

Chaos: An Interdisciplinary Journal of Nonlinear Science, Feb 2021

arXiv Bib HTML

@article{thiem2021global,
  author = {Thiem, Thomas N. and Kemeth, Felix P. and Bertalan, Tom and Laing, Carlo R. and Kevrekidis, Ioannis G.},
  title = {Global and local reduced models for interacting, heterogeneous agents},
  journal = {Chaos: An Interdisciplinary Journal of Nonlinear Science},
  volume = {31},
  number = {7},
  pages = {073139},
  year = {2021},
  doi = {10.1063/5.0055840},
  url = {https://doi.org/10.1063/5.0055840},
  eprint = {https://doi.org/10.1063/5.0055840},
}

Coarse-grained and Emergent Distributed Parameter Systems from Data

Hassan Arbabi, Felix P. Kemeth, Tom Bertalan, and 1 more author

In 2021 American Control Conference (ACC), Feb 2021

arXiv Bib HTML

@inproceedings{9483122,
  author = {Arbabi, Hassan and Kemeth, Felix P. and Bertalan, Tom and Kevrekidis, Ioannis},
  booktitle = {2021 American Control Conference (ACC)},
  title = {Coarse-grained and Emergent Distributed Parameter Systems from Data},
  year = {2021},
  volume = {},
  number = {},
  pages = {4063-4068},
  doi = {10.23919/ACC50511.2021.9483122},
  url = {https://doi.org/10.23919/ACC50511.2021.9483122},
}

2020

Combining Scatter Transform and Deep Neural Networks for Multilabel Electrocardiogram Signal Classification

Maximilian P. Oppelt, Maximilian Riehl, Felix P. Kemeth, and 1 more author

In 2020 Computing in Cardiology, Feb 2020

arXiv Bib HTML

@inproceedings{9344361,
  author = {Oppelt, Maximilian P. and Riehl, Maximilian and Kemeth, Felix P. and Steffan, Jan},
  booktitle = {2020 Computing in Cardiology},
  title = {Combining Scatter Transform and Deep Neural Networks for
                    Multilabel Electrocardiogram Signal Classification},
  year = {2020},
  volume = {},
  number = {},
  pages = {1-4},
  doi = {10.22489/CinC.2020.133},
  url = {https://doi.org/10.22489/CinC.2020.133},
}

The Effect of Data Augmentation on Classification of Atrial Fibrillation in Short Single-Lead ECG Signals Using Deep Neural Networks

Faezeh Nejati Hatamian, Nishant Ravikumar, Sulaiman Vesal, and 3 more authors

In ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), May 2020

arXiv Bib HTML

@inproceedings{hatamian20_effec_data_augmen_class_atrial,
  author = {Hatamian, Faezeh Nejati and Ravikumar, Nishant and Vesal, Sulaiman and Kemeth, Felix P. and Struck, Matthias and Maier, Andreas},
  title = {The Effect of Data Augmentation on Classification of
                    Atrial Fibrillation in Short Single-Lead ECG Signals
                    Using Deep Neural Networks},
  booktitle = {ICASSP 2020 - 2020 IEEE International Conference on
                    Acoustics, Speech and Signal Processing (ICASSP)},
  year = {2020},
  pages = {},
  doi = {10.1109/icassp40776.2020.9053800},
  url = {https://doi.org/10.1109/icassp40776.2020.9053800},
  date_added = {Fri Jan 15 14:29:33 2021},
  month = may,
}

NeurIPS 2019 Disentanglement Challenge: Improved Disentanglement through Learned Aggregation of Convolutional Feature Maps

Maximilian Seitzer, Andreas Foltyn, and Felix P. Kemeth

May 2020

arXiv Bib

@misc{seitzer2020neurips,
  title = {NeurIPS 2019 Disentanglement Challenge: Improved Disentanglement through
                    Learned Aggregation of Convolutional Feature Maps},
  author = {Seitzer, Maximilian and Foltyn, Andreas and Kemeth, Felix P.},
  year = {2020},
  eprint = {2002.12356},
  archiveprefix = {arXiv},
  primaryclass = {cs.LG},
}

2019

Cluster Singularity: the Unfolding of Clustering Behavior in Globally Coupled Stuart-Landau Oscillators

Felix P. Kemeth, Sindre W. Haugland, and Katharina Krischer

Chaos: An Interdisciplinary Journal of Nonlinear Science, May 2019

arXiv Bib HTML

@article{kemeth19_clust_singul,
  author = {Kemeth, Felix P. and Haugland, Sindre W. and Krischer, Katharina},
  title = {Cluster Singularity: the Unfolding of Clustering
                    Behavior in Globally Coupled {S}tuart-{L}andau
                    Oscillators},
  journal = {Chaos: An Interdisciplinary Journal of Nonlinear
                    Science},
  volume = {29},
  number = {2},
  pages = {023107},
  year = {2019},
  doi = {10.1063/1.5055839},
  url = {https://doi.org/10.1063/1.5055839},
  date_added = {Thu Jan 16 22:00:48 2020},
}

Lyapunov Spectra and Collective Modes of Chimera States in Globally Coupled Stuart-Landau Oscillators

Kevin Höhlein, Felix P. Kemeth, and Katharina Krischer

Physical Review E, May 2019

arXiv Bib HTML

@article{hoehlein19_lyapun_spect_collec_modes_chimer,
  author = {H{\"o}hlein, Kevin and Kemeth, Felix P. and Krischer, Katharina},
  title = {Lyapunov Spectra and Collective Modes of Chimera
                    States in Globally Coupled Stuart-Landau
                    Oscillators},
  journal = {Physical Review E},
  volume = {100},
  number = {2},
  pages = {022217},
  year = {2019},
  doi = {10.1103/physreve.100.022217},
  url = {https://doi.org/10.1103/physreve.100.022217},
  date_added = {Fri Jan 15 15:20:19 2021},
}

Symmetry Breaking in Networks of Globally Coupled Oscillators: From Clustering to Chimera States

Felix P. Kemeth

Technische Universität München, Apr 2019

Bib

@phdthesis{fpkemeth19_thesis,
  author = {Kemeth, Felix P.},
  title = {Symmetry Breaking in Networks of Globally Coupled Oscillators:
                    {F}rom Clustering to Chimera States},
  school = {Technische Universit\"{a}t M\"{u}nchen},
  year = {2019},
  address = {Garching},
  month = apr,
}

2018

Symmetries of Chimera States

Felix P. Kemeth, Sindre W. Haugland, and Katharina Krischer

Physical Review Letters, Apr 2018

Abs arXiv Bib HTML

Symmetry broken states arise naturally in oscillatory networks. In this Letter, we investigate chaotic attractors in an ensemble of four mean-coupled Stuart-Landau oscillators with two oscillators being synchronized. We report that these states with partially broken symmetry, so-called chimera states, have different setwise symmetries in the incoherent oscillators, and in particular, some are and some are not invariant under a permutation symmetry on average. This allows for a classification of different chimera states in small networks. We conclude our report with a discussion of related states in spatially extended systems, which seem to inherit the symmetry properties of their counterparts in small networks.
@article{kemeth18_symmet_chimer_states, author = {Kemeth, Felix P. and Haugland, Sindre W. and Krischer, Katharina}, title = {Symmetries of Chimera States}, journal = {Physical Review Letters}, volume = {120}, number = {21}, pages = {214101}, year = {2018}, doi = {10.1103/physrevlett.120.214101}, url = {https://doi.org/10.1103/physrevlett.120.214101}, date_added = {Fri Jan 15 15:21:46 2021}, dimensions = {true}, }

An Emergent Space for Distributed Data With Hidden Internal Order Through Manifold Learning

Felix P. Kemeth, Sindre W. Haugland, Felix Dietrich, and 7 more authors

IEEE Access, Apr 2018

arXiv Bib HTML

@article{kemeth18_emerg_space_distr_data_with,
  author = {Kemeth, Felix P. and Haugland, Sindre W. and Dietrich, Felix and Bertalan, Tom and Hohlein, Kevin and Li, Qianxiao and Bollt, Erik M. and Talmon, Ronen and Krischer, Katharina and Kevrekidis, Ioannis G.},
  title = {An Emergent Space for Distributed Data With Hidden
                    Internal Order Through Manifold Learning},
  journal = {IEEE Access},
  volume = {6},
  number = {},
  pages = {77402-77413},
  year = {2018},
  doi = {10.1109/access.2018.2882777},
  url = {https://doi.org/10.1109/access.2018.2882777},
  date_added = {Wed Apr 29 11:08:04 2020},
}

2016

A Classification Scheme for Chimera States

Felix P. Kemeth, Sindre W. Haugland, Lennart Schmidt, and 2 more authors

Chaos: An Interdisciplinary Journal of Nonlinear Science, Apr 2016

Abs arXiv Bib HTML Code

We present a universal characterization scheme for chimera states applicable to both numerical and experimental data sets. The scheme is based on two correlation measures that enable a meaningful definition of chimera states as well as their classification into three categories: stationary, turbulent, and breathing. In addition, these categories can be further subdivided according to the time-stationarity of these two measures. We demonstrate that this approach is both consistent with previously recognized chimera states and enables us to classify states as chimeras which have not been categorized as such before. Furthermore, the scheme allows for a qualitative and quantitative comparison of experimental chimeras with chimeras obtained through numerical simulations.
@article{kemeth16_class_schem_chimer_states, author = {Kemeth, Felix P. and Haugland, Sindre W. and Schmidt, Lennart and Kevrekidis, Ioannis G. and Krischer, Katharina}, title = {A Classification Scheme for Chimera States}, journal = {Chaos: An Interdisciplinary Journal of Nonlinear Science}, volume = {26}, number = {9}, pages = {094815}, year = {2016}, doi = {10.1063/1.4959804}, url = {https://doi.org/10.1063/1.4959804}, date_added = {Fri Jan 15 15:16:41 2021}, dimensions = {true}, }

2014

A Capacitance Mediated Positive Differential Resistance Oscillator Model for Electrochemical Systems Involving a Surface Layer

Carla Zensen, Konrad Schönleber, Felix P. Kemeth, and 1 more author

The Journal of Physical Chemistry C, Apr 2014

Bib HTML

@article{zensen14_capac_mediat_posit_differ_resis,
  author = {Zensen, Carla and Sch{\"o}nleber, Konrad and Kemeth, Felix P. and Krischer, Katharina},
  title = {A Capacitance Mediated Positive Differential
                    Resistance Oscillator Model for Electrochemical
                    Systems Involving a Surface Layer},
  journal = {The Journal of Physical Chemistry C},
  volume = {118},
  number = {42},
  pages = {24407-24414},
  year = {2014},
  doi = {10.1021/jp505418x},
  url = {https://doi.org/10.1021/jp505418x},
  date_added = {Tue Dec 15 11:23:45 2020},
}