Contrastive learning of strong-mixing continuous-time stochastic processes

Bingbin Liu, Pradeep Ravikumar, Andrej Risteski
Proceedings of The 24th International Conference on Artificial Intelligence and Statistics, PMLR 130:3151-3159, 2021.

Abstract

Contrastive learning is a family of self-supervised methods where a model is trained to solve a classification task constructed from unlabeled data. It has recently emerged as one of the leading learning paradigms in the absence of labels across many different domains (e.g. brain imaging, text, images). However, theoretical understanding of many aspects of training, both statistical and algorithmic, remain fairly elusive. In this work, we study the setting of time series—more precisely, when we get data from a strong-mixing continuous-time stochastic process. We show that a properly constructed contrastive learning task can be used to the transition kernel for small-to-mid-range intervals in the diffusion case. Moreover, we give sample complexity bounds for solving this task and quantitatively characterize what the value of the contrastive loss implies for distributional closeness of the learned kernel. As a byproduct, we illuminate the appropriate settings for the contrastive distribution, as well as other hyperparameters in this setup.

Cite this Paper


BibTeX
@InProceedings{pmlr-v130-liu21h, title = { Contrastive learning of strong-mixing continuous-time stochastic processes }, author = {Liu, Bingbin and Ravikumar, Pradeep and Risteski, Andrej}, booktitle = {Proceedings of The 24th International Conference on Artificial Intelligence and Statistics}, pages = {3151--3159}, year = {2021}, editor = {Banerjee, Arindam and Fukumizu, Kenji}, volume = {130}, series = {Proceedings of Machine Learning Research}, month = {13--15 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v130/liu21h/liu21h.pdf}, url = {https://proceedings.mlr.press/v130/liu21h.html}, abstract = { Contrastive learning is a family of self-supervised methods where a model is trained to solve a classification task constructed from unlabeled data. It has recently emerged as one of the leading learning paradigms in the absence of labels across many different domains (e.g. brain imaging, text, images). However, theoretical understanding of many aspects of training, both statistical and algorithmic, remain fairly elusive. In this work, we study the setting of time series—more precisely, when we get data from a strong-mixing continuous-time stochastic process. We show that a properly constructed contrastive learning task can be used to the transition kernel for small-to-mid-range intervals in the diffusion case. Moreover, we give sample complexity bounds for solving this task and quantitatively characterize what the value of the contrastive loss implies for distributional closeness of the learned kernel. As a byproduct, we illuminate the appropriate settings for the contrastive distribution, as well as other hyperparameters in this setup. } }
Endnote
%0 Conference Paper %T Contrastive learning of strong-mixing continuous-time stochastic processes %A Bingbin Liu %A Pradeep Ravikumar %A Andrej Risteski %B Proceedings of The 24th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2021 %E Arindam Banerjee %E Kenji Fukumizu %F pmlr-v130-liu21h %I PMLR %P 3151--3159 %U https://proceedings.mlr.press/v130/liu21h.html %V 130 %X Contrastive learning is a family of self-supervised methods where a model is trained to solve a classification task constructed from unlabeled data. It has recently emerged as one of the leading learning paradigms in the absence of labels across many different domains (e.g. brain imaging, text, images). However, theoretical understanding of many aspects of training, both statistical and algorithmic, remain fairly elusive. In this work, we study the setting of time series—more precisely, when we get data from a strong-mixing continuous-time stochastic process. We show that a properly constructed contrastive learning task can be used to the transition kernel for small-to-mid-range intervals in the diffusion case. Moreover, we give sample complexity bounds for solving this task and quantitatively characterize what the value of the contrastive loss implies for distributional closeness of the learned kernel. As a byproduct, we illuminate the appropriate settings for the contrastive distribution, as well as other hyperparameters in this setup.
APA
Liu, B., Ravikumar, P. & Risteski, A.. (2021). Contrastive learning of strong-mixing continuous-time stochastic processes . Proceedings of The 24th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 130:3151-3159 Available from https://proceedings.mlr.press/v130/liu21h.html.

Related Material