Karakulev A; Zachariah D; Singh P
Adaptive Parameter-Free Robust Learning using Latent Bernoulli Variables Conference
arXiv preprint, vol. arXiv:2312.00585, 2023.
@conference{nokey,
title = {Adaptive Parameter-Free Robust Learning using Latent Bernoulli Variables},
author = {Aleksandr Karakulev and Dave Zachariah and Prashant Singh},
url = {https://arxiv.org/abs/2312.00585},
doi = {https://doi.org/10.48550/arXiv.2312.00585},
year = {2023},
date = {2023-12-01},
urldate = {2023-12-01},
booktitle = {arXiv preprint},
volume = {arXiv:2312.00585},
abstract = {We present an efficient parameter-free approach for statistical learning from corrupted training sets. We identify corrupted and non-corrupted samples using latent Bernoulli variables, and therefore formulate the robust learning problem as maximization of the likelihood where latent variables are marginalized out. The resulting optimization problem is solved via variational inference using an efficient Expectation-Maximization based method. The proposed approach improves over the state-of-the-art by automatically inferring the corruption level and identifying outliers, while adding minimal computational overhead. We demonstrate our robust learning method on a wide variety of machine learning tasks including online learning and deep learning where it exhibits ability to adapt to different levels of noise and attain high prediction accuracy.},
keywords = {Bayesian Inference, Deep Learning, Optimization, Robust Learning},
pubstate = {published},
tppubtype = {conference}
}
We present an efficient parameter-free approach for statistical learning from corrupted training sets. We identify corrupted and non-corrupted samples using latent Bernoulli variables, and therefore formulate the robust learning problem as maximization of the likelihood where latent variables are marginalized out. The resulting optimization problem is solved via variational inference using an efficient Expectation-Maximization based method. The proposed approach improves over the state-of-the-art by automatically inferring the corruption level and identifying outliers, while adding minimal computational overhead. We demonstrate our robust learning method on a wide variety of machine learning tasks including online learning and deep learning where it exhibits ability to adapt to different levels of noise and attain high prediction accuracy.
Singh P; Wrede F; Hellander A
Scalable machine learning-assisted model exploration and inference using Sciope Journal Article
In: Bioinformatics, vol. 37, no. 2, pp. 279–281, 2021.
@article{singh2021scalable,
title = {Scalable machine learning-assisted model exploration and inference using Sciope},
author = {Prashant Singh and Fredrik Wrede and Andreas Hellander},
url = {https://academic.oup.com/bioinformatics/article/37/2/279/5876021},
doi = {https://doi.org/10.1093/bioinformatics/btaa673},
year = {2021},
date = {2021-01-01},
urldate = {2021-01-01},
journal = {Bioinformatics},
volume = {37},
number = {2},
pages = {279--281},
publisher = {Oxford University Press},
keywords = {Bayesian Inference, Deep Learning, Inverse Problem, Optimization, Software, Surrogate Modeling},
pubstate = {published},
tppubtype = {article}
}
Singh P; Vats E; Hast A
Learning surrogate models of document image quality metrics for automated document image processing Proceedings Article
In: 2018 13th IAPR International Workshop on Document Analysis Systems (DAS), pp. 67–72, IEEE 2018.
@inproceedings{singh2018learning,
title = {Learning surrogate models of document image quality metrics for automated document image processing},
author = {Prashant Singh and Ekta Vats and Anders Hast},
url = {https://ieeexplore.ieee.org/abstract/document/8395173
https://arxiv.org/pdf/1712.03738},
doi = {https://doi.org/10.1109/DAS.2018.14},
year = {2018},
date = {2018-01-01},
urldate = {2018-01-01},
booktitle = {2018 13th IAPR International Workshop on Document Analysis Systems (DAS)},
pages = {67--72},
organization = {IEEE},
keywords = {Optimization, Surrogate Modeling},
pubstate = {published},
tppubtype = {inproceedings}
}
Singh P; Hellander A
Hyperparameter optimization for approximate Bayesian computation Proceedings Article
In: 2018 Winter Simulation Conference (WSC), pp. 1718–1729, IEEE 2018.
@inproceedings{singh2018hyperparameter,
title = {Hyperparameter optimization for approximate Bayesian computation},
author = {Prashant Singh and Andreas Hellander},
url = {https://ieeexplore.ieee.org/abstract/document/8632304
},
doi = {https://doi.org/10.1109/WSC.2018.8632304},
year = {2018},
date = {2018-01-01},
urldate = {2018-01-01},
booktitle = {2018 Winter Simulation Conference (WSC)},
pages = {1718--1729},
organization = {IEEE},
keywords = {Bayesian Inference, Inverse Problem, Optimization},
pubstate = {published},
tppubtype = {inproceedings}
}
Singh P; Hellander A
Multi-objective optimization driven construction of uniform priors for likelihood-free parameter inference Proceedings Article
In: ESM 2018, October 24--26, Ghent, Belgium, pp. 22–27, EUROSIS 2018.
@inproceedings{singh2018multi,
title = {Multi-objective optimization driven construction of uniform priors for likelihood-free parameter inference},
author = {Prashant Singh and Andreas Hellander},
year = {2018},
date = {2018-01-01},
urldate = {2018-01-01},
booktitle = {ESM 2018, October 24--26, Ghent, Belgium},
pages = {22--27},
organization = {EUROSIS},
keywords = {Bayesian Inference, Inverse Problem, Optimization},
pubstate = {published},
tppubtype = {inproceedings}
}
Singh P; van der Herten J; Deschrijver D; Couckuyt I; Dhaene T
A sequential sampling strategy for adaptive classification of computationally expensive data Journal Article
In: Structural and Multidisciplinary Optimization, vol. 55, no. 4, pp. 1425–1438, 2017.
@article{singh2017sequential,
title = {A sequential sampling strategy for adaptive classification of computationally expensive data},
author = {Prashant Singh and Joachim van der Herten and Dirk Deschrijver and Ivo Couckuyt and Tom Dhaene},
url = {https://link.springer.com/article/10.1007/s00158-016-1584-1
https://users.ugent.be/~didschri/papers/2017_04_Springer_SMO.pdf
},
doi = {https://doi.org/10.1007/s00158-016-1584-1},
year = {2017},
date = {2017-01-01},
urldate = {2017-01-01},
journal = {Structural and Multidisciplinary Optimization},
volume = {55},
number = {4},
pages = {1425--1438},
publisher = {Springer Berlin Heidelberg},
keywords = {Classification, Inverse Problem, Optimization, Sampling, Surrogate Modeling},
pubstate = {published},
tppubtype = {article}
}
Singh P; Rossi M; Couckuyt I; Deschrijver D; Rogier H; Dhaene T
Constrained multi-objective antenna design optimization using surrogates Journal Article
In: International Journal of Numerical Modelling: Electronic Networks, Devices and Fields, vol. 30, no. 6, pp. e2248, 2017.
@article{singh2017constrained,
title = {Constrained multi-objective antenna design optimization using surrogates},
author = {Prashant Singh and Marco Rossi and Ivo Couckuyt and Dirk Deschrijver and Hendrik Rogier and Tom Dhaene},
url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/jnm.2248
http://sumo.intec.ugent.be/sites/default/files/dirk_pubs/2017_11_Wiley_IJNM.pdf},
doi = {https://doi.org/10.1002/jnm.2248},
year = {2017},
date = {2017-01-01},
urldate = {2017-01-01},
journal = {International Journal of Numerical Modelling: Electronic Networks, Devices and Fields},
volume = {30},
number = {6},
pages = {e2248},
keywords = {Optimization, Surrogate Modeling},
pubstate = {published},
tppubtype = {article}
}
Singh P; Couckuyt I; Elsayed K; Deschrijver D; Dhaene T
Multi-objective geometry optimization of a gas cyclone using triple-fidelity co-kriging surrogate models Journal Article
In: Journal of Optimization Theory and Applications, vol. 175, no. 1, pp. 172–193, 2017.
@article{singh2017multi,
title = {Multi-objective geometry optimization of a gas cyclone using triple-fidelity co-kriging surrogate models},
author = {Prashant Singh and Ivo Couckuyt and Khairy Elsayed and Dirk Deschrijver and Tom Dhaene},
url = {https://link.springer.com/article/10.1007/s10957-017-1114-3
http://sumo.intec.ugent.be/sites/default/files/dirk_pubs/2017_10_Springer_JOTA.pdf},
doi = {https://doi.org/10.1007/s10957-017-1114-3},
year = {2017},
date = {2017-01-01},
urldate = {2017-01-01},
journal = {Journal of Optimization Theory and Applications},
volume = {175},
number = {1},
pages = {172--193},
publisher = {Springer US},
keywords = {Optimization, Surrogate Modeling},
pubstate = {published},
tppubtype = {article}
}
Vats E; Hast A; Singh P
Automatic document image binarization using bayesian optimization Proceedings Article
In: Proceedings of the 4th International Workshop on Historical Document Imaging and Processing, pp. 89–94, 2017.
@inproceedings{vats2017automatic,
title = {Automatic document image binarization using bayesian optimization},
author = {Ekta Vats and Anders Hast and Prashant Singh},
url = {https://dl.acm.org/doi/abs/10.1145/3151509.3151520
https://arxiv.org/pdf/1709.01782
},
doi = {https://doi.org/10.1145/3151509.3151520},
year = {2017},
date = {2017-01-01},
urldate = {2017-01-01},
booktitle = {Proceedings of the 4th International Workshop on Historical Document Imaging and Processing},
pages = {89--94},
keywords = {Optimization, Surrogate Modeling},
pubstate = {published},
tppubtype = {inproceedings}
}
Singh P; Couckuyt I; Elsayed K; Deschrijver D; Dhaene T
Shape optimization of a cyclone separator using multi-objective surrogate-based optimization Journal Article
In: Applied Mathematical Modelling, vol. 40, no. 5-6, pp. 4248–4259, 2016.
@article{singh2016shape,
title = {Shape optimization of a cyclone separator using multi-objective surrogate-based optimization},
author = {Prashant Singh and Ivo Couckuyt and Khairy Elsayed and Dirk Deschrijver and Tom Dhaene},
url = {https://www.sciencedirect.com/science/article/pii/S0307904X15007210
http://sumo.intec.ugent.be/sites/default/files/dirk_pubs/2016_03_Elsevier_APM.pdf},
doi = {https://doi.org/10.1016/j.apm.2015.11.007},
year = {2016},
date = {2016-01-01},
urldate = {2016-01-01},
journal = {Applied Mathematical Modelling},
volume = {40},
number = {5-6},
pages = {4248--4259},
publisher = {Elsevier},
keywords = {Optimization, Sampling, Surrogate Modeling},
pubstate = {published},
tppubtype = {article}
}
Singh P
Design of experiments for model-based optimization PhD Thesis
Ghent University, 2016.
@phdthesis{singh2016design,
title = {Design of experiments for model-based optimization},
author = {Prashant Singh},
url = {https://biblio.ugent.be/publication/7223691/file/7223692},
year = {2016},
date = {2016-01-01},
urldate = {2016-01-01},
school = {Ghent University},
keywords = {Optimization, Sampling, Surrogate Modeling},
pubstate = {published},
tppubtype = {phdthesis}
}
Singh P; Claeys T; Vandenbosch G A; Pissoort D
Automated line-based sequential sampling and modeling algorithm for EMC near-field scanning Journal Article
In: IEEE Transactions on Electromagnetic Compatibility, vol. 59, no. 2, pp. 704–709, 2016.
@article{singh2016automated,
title = {Automated line-based sequential sampling and modeling algorithm for EMC near-field scanning},
author = {Prashant Singh and Tim Claeys and Guy AE Vandenbosch and Davy Pissoort},
url = {https://ieeexplore.ieee.org/abstract/document/7776759
https://lirias.kuleuven.be/retrieve/639699},
doi = {https://doi.org/10.1109/TEMC.2016.2632741},
year = {2016},
date = {2016-01-01},
urldate = {2016-01-01},
journal = {IEEE Transactions on Electromagnetic Compatibility},
volume = {59},
number = {2},
pages = {704--709},
publisher = {IEEE},
keywords = {Optimization, Sampling, Surrogate Modeling},
pubstate = {published},
tppubtype = {article}
}
Gong X; Trogh J; Braet Q; Tanghe E; Singh P; Plets D; Hoebeke J; Deschrijver D; Dhaene T; Martens L; others
Measurement-based wireless network planning, monitoring, and reconfiguration solution for robust radio communications in indoor factories Journal Article
In: IET Science, Measurement & Technology, vol. 10, no. 4, pp. 375–382, 2016.
@article{gong2016measurement,
title = {Measurement-based wireless network planning, monitoring, and reconfiguration solution for robust radio communications in indoor factories},
author = {Xu Gong and Jens Trogh and Quentin Braet and Emmeric Tanghe and Prashant Singh and David Plets and Jeroen Hoebeke and Dirk Deschrijver and Tom Dhaene and Luc Martens and others},
url = {https://ietresearch.onlinelibrary.wiley.com/doi/full/10.1049/iet-smt.2015.0213
https://ietresearch.onlinelibrary.wiley.com/doi/pdfdirect/10.1049/iet-smt.2015.0213},
doi = {https://doi.org/10.1049/iet-smt.2015.0213},
year = {2016},
date = {2016-01-01},
urldate = {2016-01-01},
journal = {IET Science, Measurement & Technology},
volume = {10},
number = {4},
pages = {375--382},
publisher = {The Institution of Engineering and Technology},
keywords = {Optimization, Sampling, Surrogate Modeling},
pubstate = {published},
tppubtype = {article}
}
Vermeeren G; Singh P; Aerts S; Deschrijver D; Dhaene T; Joseph W; Martens L
Surrogate-based fast peak mass-averaged SAR assessment Proceedings Article
In: Annual Meeting of the Bioelectromagnetics Society and the European BioElectromagnetics Association (BioEM 2015), pp. 135–137, 2015.
@inproceedings{vermeeren2015surrogate,
title = {Surrogate-based fast peak mass-averaged SAR assessment},
author = {Günter Vermeeren and Prashant Singh and Sam Aerts and Dirk Deschrijver and Tom Dhaene and Wout Joseph and Luc Martens},
url = {https://biblio.ugent.be/publication/7257304/file/7257373.pdf},
year = {2015},
date = {2015-01-01},
urldate = {2015-01-01},
booktitle = {Annual Meeting of the Bioelectromagnetics Society and the European BioElectromagnetics Association (BioEM 2015)},
pages = {135--137},
keywords = {Inverse Problem, Optimization, Sampling, Surrogate Modeling},
pubstate = {published},
tppubtype = {inproceedings}
}
Singh P; Couckuyt I; Ferranti F; Dhaene T
A constrained multi-objective surrogate-based optimization algorithm Proceedings Article
In: 2014 IEEE Congress on Evolutionary Computation (CEC), pp. 3080–3087, IEEE 2014.
@inproceedings{singh2014constrained,
title = {A constrained multi-objective surrogate-based optimization algorithm},
author = {Prashant Singh and Ivo Couckuyt and Francesco Ferranti and Tom Dhaene},
url = {https://ieeexplore.ieee.org/abstract/document/6900581
https://biblio.ugent.be/publication/5733039/file/5733050.pdf},
doi = {https://doi.org/10.1109/CEC.2014.6900581},
year = {2014},
date = {2014-01-01},
urldate = {2014-01-01},
booktitle = {2014 IEEE Congress on Evolutionary Computation (CEC)},
pages = {3080--3087},
organization = {IEEE},
keywords = {Optimization, Sampling, Surrogate Modeling},
pubstate = {published},
tppubtype = {inproceedings}
}
Singh P; Ferranti F; Deschrijver D; Couckuyt I; Dhaene T
Classification aided domain reduction for high dimensional optimization Proceedings Article
In: Proceedings of the Winter Simulation Conference 2014, pp. 3928–3939, IEEE 2014.
@inproceedings{singh2014classification,
title = {Classification aided domain reduction for high dimensional optimization},
author = {Prashant Singh and Francesco Ferranti and Dirk Deschrijver and Ivo Couckuyt and Tom Dhaene},
url = {https://ieeexplore.ieee.org/abstract/document/7020218
https://biblio.ugent.be/publication/5955500/file/5955522},
doi = {https://doi.org/10.1109/WSC.2014.7020218},
year = {2014},
date = {2014-01-01},
urldate = {2014-01-01},
booktitle = {Proceedings of the Winter Simulation Conference 2014},
pages = {3928--3939},
organization = {IEEE},
keywords = {Classification, Optimization, Sampling, Surrogate Modeling},
pubstate = {published},
tppubtype = {inproceedings}
}
Singh P; Deschrijver D; Pissoort D; Dhaene T
Adaptive classification algorithm for EMC-compliance testing of electronic devices Journal Article
In: Electronics Letters, vol. 49, no. 24, pp. 1526–1528, 2013.
@article{singh2013adaptive,
title = {Adaptive classification algorithm for EMC-compliance testing of electronic devices},
author = {Prashant Singh and Dirk Deschrijver and Davy Pissoort and Tom Dhaene},
url = {https://ietresearch.onlinelibrary.wiley.com/doi/full/10.1049/el.2013.2766
https://ietresearch.onlinelibrary.wiley.com/doi/epdf/10.1049/el.2013.2766},
doi = {https://doi.org/10.1049/el.2013.2766},
year = {2013},
date = {2013-01-01},
urldate = {2013-01-01},
journal = {Electronics Letters},
volume = {49},
number = {24},
pages = {1526--1528},
publisher = {The Institution of Engineering and Technology},
keywords = {Classification, Inverse Problem, Optimization, Surrogate Modeling},
pubstate = {published},
tppubtype = {article}
}
Singh P; Deschrijver D; Dhaene T
A balanced sequential design strategy for global surrogate modeling Proceedings Article
In: 2013 Winter Simulations Conference (WSC), pp. 2172–2179, IEEE 2013.
@inproceedings{singh2013balanced,
title = {A balanced sequential design strategy for global surrogate modeling},
author = {Prashant Singh and Dirk Deschrijver and Tom Dhaene},
url = {https://ieeexplore.ieee.org/abstract/document/6721594
https://biblio.ugent.be/publication/4315532/file/4315539},
doi = {https://doi.org/10.1109/WSC.2013.6721594},
year = {2013},
date = {2013-01-01},
urldate = {2013-01-01},
booktitle = {2013 Winter Simulations Conference (WSC)},
pages = {2172--2179},
organization = {IEEE},
keywords = {Optimization, Sampling, Surrogate Modeling},
pubstate = {published},
tppubtype = {inproceedings}
}
Singh P; Deschrijver D; Pissoort D; Dhaene T
Accurate hotspot localization by sampling the near-field pattern of electronic devices Journal Article
In: IEEE Transactions on Electromagnetic Compatibility, vol. 55, no. 6, pp. 1365–1368, 2013.
@article{singh2013accurate,
title = {Accurate hotspot localization by sampling the near-field pattern of electronic devices},
author = {Prashant Singh and Dirk Deschrijver and Davy Pissoort and Tom Dhaene},
url = {https://ieeexplore.ieee.org/abstract/document/6522868
https://biblio.ugent.be/publication/4210867/file/4210870.pdf},
doi = {https://doi.org/10.1109/TEMC.2013.2265158},
year = {2013},
date = {2013-01-01},
urldate = {2013-01-01},
journal = {IEEE Transactions on Electromagnetic Compatibility},
volume = {55},
number = {6},
pages = {1365--1368},
publisher = {IEEE},
keywords = {Inverse Problem, Optimization, Sampling},
pubstate = {published},
tppubtype = {article}
}
Singh P; Deschrijver D; Pissoort D; Dhaene T
Efficient measurement procedure for hotspot detection in near-field pattern of electronic devices Proceedings Article
In: BESTCOM Meeting, 2013.
@inproceedings{singh2013efficient,
title = {Efficient measurement procedure for hotspot detection in near-field pattern of electronic devices},
author = {Prashant Singh and Dirk Deschrijver and Davy Pissoort and Tom Dhaene},
url = {https://biblio.ugent.be/publication/4083180/file/4083328.pdf},
year = {2013},
date = {2013-01-01},
urldate = {2013-01-01},
booktitle = {BESTCOM Meeting},
keywords = {Inverse Problem, Optimization, Sampling, Surrogate Modeling},
pubstate = {published},
tppubtype = {inproceedings}
}