Toward Learning Distributions of Distributions

Moritz Wohlstein, Ulf Brefeld
Proceedings of the 6th Northern Lights Deep Learning Conference (NLDL), PMLR 265:269-275, 2025.

Abstract

We propose a novel generative deep learning architecture based on generative moment matching networks. The objective of our model is to learn a distribution over distributions and generate new sample distributions following the (possibly complex) distribution of training data. We derive a custom loss function for our model based on the maximum mean discrepancy test. Our model is evaluated on different datasets where we investigate the influence of hyperparameters on performance.

Cite this Paper

BibTeX
@InProceedings{pmlr-v265-wohlstein25a, title = {Toward Learning Distributions of Distributions}, author = {Wohlstein, Moritz and Brefeld, Ulf}, booktitle = {Proceedings of the 6th Northern Lights Deep Learning Conference (NLDL)}, pages = {269--275}, year = {2025}, editor = {Lutchyn, Tetiana and Ramírez Rivera, Adín and Ricaud, Benjamin}, volume = {265}, series = {Proceedings of Machine Learning Research}, month = {07--09 Jan}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v265/main/assets/wohlstein25a/wohlstein25a.pdf}, url = {https://proceedings.mlr.press/v265/wohlstein25a.html}, abstract = {We propose a novel generative deep learning architecture based on generative moment matching networks. The objective of our model is to learn a distribution over distributions and generate new sample distributions following the (possibly complex) distribution of training data. We derive a custom loss function for our model based on the maximum mean discrepancy test. Our model is evaluated on different datasets where we investigate the influence of hyperparameters on performance.} }
Endnote
%0 Conference Paper %T Toward Learning Distributions of Distributions %A Moritz Wohlstein %A Ulf Brefeld %B Proceedings of the 6th Northern Lights Deep Learning Conference (NLDL) %C Proceedings of Machine Learning Research %D 2025 %E Tetiana Lutchyn %E Adín Ramírez Rivera %E Benjamin Ricaud %F pmlr-v265-wohlstein25a %I PMLR %P 269--275 %U https://proceedings.mlr.press/v265/wohlstein25a.html %V 265 %X We propose a novel generative deep learning architecture based on generative moment matching networks. The objective of our model is to learn a distribution over distributions and generate new sample distributions following the (possibly complex) distribution of training data. We derive a custom loss function for our model based on the maximum mean discrepancy test. Our model is evaluated on different datasets where we investigate the influence of hyperparameters on performance.
APA
Wohlstein, M. & Brefeld, U.. (2025). Toward Learning Distributions of Distributions. Proceedings of the 6th Northern Lights Deep Learning Conference (NLDL), in Proceedings of Machine Learning Research 265:269-275 Available from https://proceedings.mlr.press/v265/wohlstein25a.html.

Related Material