DisCoV: Disentangling Time Series Representations via Contrastive based $l$-Variational Inference

Khalid Oublal, Said Ladjal, David Benhaiem, Emmanuel Le-borgne, François Roueff
Proceedings of UniReps: the First Workshop on Unifying Representations in Neural Models, PMLR 243:223-236, 2024.

Abstract

Learning disentangled representations is crucial for Time Series, offering benefits like feature derivation and improved interpretability, thereby enhancing task performance. We focus on disentangled representation learning for home appliance electricity usage, enabling users to understand and optimize their consumption for a reduced carbon footprint. Our approach frames the problem as disentangling each attribute’s role in total consumption (e.g., dishwashers, fridges, …). Unlike existing methods assuming attribute independence, we acknowledge real-world time series attribute correlations, like the operating of dishwashers and washing machines during the winter season. To tackle this, we employ weakly supervised contrastive disentanglement, facilitating representation generalization across diverse correlated scenarios and new households. Our method utilizes innovative $l$-variational inference layers with self-attention, effectively addressing temporal dependencies across bottom-up and top-down networks. We find that DisCoV (Disentangling via Contrastive $l$-Variational) can enhance the task of reconstructing electricity consumption for individual appliances. We introduce TDS (Time Disentangling Score) to gauge disentanglement quality. TDS reliably reflects disentanglement performance, making it a valuable metric for evaluating time series representations. Code available at https://anonymous.4open.science/r/DisCo.

Cite this Paper

BibTeX
@InProceedings{pmlr-v243-oublal24a, title = {DisCoV: Disentangling Time Series Representations via Contrastive based $l$-Variational Inference}, author = {Oublal, Khalid and Ladjal, Said and Benhaiem, David and Le-borgne, Emmanuel and Roueff, Fran\c{c}ois}, booktitle = {Proceedings of UniReps: the First Workshop on Unifying Representations in Neural Models}, pages = {223--236}, year = {2024}, editor = {Fumero, Marco and Rodolá, Emanuele and Domine, Clementine and Locatello, Francesco and Dziugaite, Karolina and Mathilde, Caron}, volume = {243}, series = {Proceedings of Machine Learning Research}, month = {15 Dec}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v243/oublal24a/oublal24a.pdf}, url = {https://proceedings.mlr.press/v243/oublal24a.html}, abstract = {Learning disentangled representations is crucial for Time Series, offering benefits like feature derivation and improved interpretability, thereby enhancing task performance. We focus on disentangled representation learning for home appliance electricity usage, enabling users to understand and optimize their consumption for a reduced carbon footprint. Our approach frames the problem as disentangling each attribute’s role in total consumption (e.g., dishwashers, fridges, …). Unlike existing methods assuming attribute independence, we acknowledge real-world time series attribute correlations, like the operating of dishwashers and washing machines during the winter season. To tackle this, we employ weakly supervised contrastive disentanglement, facilitating representation generalization across diverse correlated scenarios and new households. Our method utilizes innovative $l$-variational inference layers with self-attention, effectively addressing temporal dependencies across bottom-up and top-down networks. We find that DisCoV (Disentangling via Contrastive $l$-Variational) can enhance the task of reconstructing electricity consumption for individual appliances. We introduce TDS (Time Disentangling Score) to gauge disentanglement quality. TDS reliably reflects disentanglement performance, making it a valuable metric for evaluating time series representations. Code available at https://anonymous.4open.science/r/DisCo.} }
Endnote
%0 Conference Paper %T DisCoV: Disentangling Time Series Representations via Contrastive based $l$-Variational Inference %A Khalid Oublal %A Said Ladjal %A David Benhaiem %A Emmanuel Le-borgne %A François Roueff %B Proceedings of UniReps: the First Workshop on Unifying Representations in Neural Models %C Proceedings of Machine Learning Research %D 2024 %E Marco Fumero %E Emanuele Rodolá %E Clementine Domine %E Francesco Locatello %E Karolina Dziugaite %E Caron Mathilde %F pmlr-v243-oublal24a %I PMLR %P 223--236 %U https://proceedings.mlr.press/v243/oublal24a.html %V 243 %X Learning disentangled representations is crucial for Time Series, offering benefits like feature derivation and improved interpretability, thereby enhancing task performance. We focus on disentangled representation learning for home appliance electricity usage, enabling users to understand and optimize their consumption for a reduced carbon footprint. Our approach frames the problem as disentangling each attribute’s role in total consumption (e.g., dishwashers, fridges, …). Unlike existing methods assuming attribute independence, we acknowledge real-world time series attribute correlations, like the operating of dishwashers and washing machines during the winter season. To tackle this, we employ weakly supervised contrastive disentanglement, facilitating representation generalization across diverse correlated scenarios and new households. Our method utilizes innovative $l$-variational inference layers with self-attention, effectively addressing temporal dependencies across bottom-up and top-down networks. We find that DisCoV (Disentangling via Contrastive $l$-Variational) can enhance the task of reconstructing electricity consumption for individual appliances. We introduce TDS (Time Disentangling Score) to gauge disentanglement quality. TDS reliably reflects disentanglement performance, making it a valuable metric for evaluating time series representations. Code available at https://anonymous.4open.science/r/DisCo.
APA
Oublal, K., Ladjal, S., Benhaiem, D., Le-borgne, E. & Roueff, F.. (2024). DisCoV: Disentangling Time Series Representations via Contrastive based $l$-Variational Inference. Proceedings of UniReps: the First Workshop on Unifying Representations in Neural Models, in Proceedings of Machine Learning Research 243:223-236 Available from https://proceedings.mlr.press/v243/oublal24a.html.

Related Material