# Deep Temporal Contrastive Clustering

arXiv: https://arxiv.org/abs/2212.14366 · HTML: https://mlsanity.com/paper/2212.14366/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 59 references that are in the map.
- First posted on arXiv 29 Dec 2022.

## Builds on (most cited first, 18 of 18)

- [Unsupervised Deep Embedding for Clustering Analysis](https://mlsanity.com/paper/1511.06335/) (2015)
- [Prototypical Contrastive Learning of Unsupervised Representations](https://mlsanity.com/paper/2005.04966/) (2020)
- [The UCR Time Series Archive](https://mlsanity.com/paper/1810.07758/) (2018)
- [Contrastive Clustering](https://mlsanity.com/paper/2009.09687/) (2020)
- [Dilated Recurrent Neural Networks](https://mlsanity.com/paper/1710.02224/) (2017)
- [Crafting Better Contrastive Views for Siamese Representation Learning](https://mlsanity.com/paper/2202.03278/) (2022)
- [Ultra-Scalable Spectral Clustering and Ensemble Clustering](https://mlsanity.com/paper/1903.01057/) (2019)
- [Deep Temporal Clustering : Fully Unsupervised Learning of Time-Domain Features](https://mlsanity.com/paper/1802.01059/) (2018)
- [Fast Multi-view Clustering via Ensembles: Towards Scalability, Superiority, and Simplicity](https://mlsanity.com/paper/2203.11572/) (2022)
- [Locally Weighted Ensemble Clustering](https://mlsanity.com/paper/1605.05011/) (2016)
- [Toward Multidiversified Ensemble Clustering of High-Dimensional Data: From Subspaces to Metrics and Beyond](https://mlsanity.com/paper/1710.03113/) (2017)
- [Multi-view Graph Learning by Joint Modeling of Consistency and Inconsistency](https://mlsanity.com/paper/2008.10208/) (2020)
- [Enhanced Ensemble Clustering via Fast Propagation of Cluster-wise Similarities](https://mlsanity.com/paper/1810.12544/) (2018)
- [Structured Graph Learning for Clustering and Semi-supervised Classification](https://mlsanity.com/paper/2008.13429/) (2020)
- [Strongly Augmented Contrastive Clustering](https://mlsanity.com/paper/2206.00380/) (2022)
- [Seeking Commonness and Inconsistencies: A Jointly Smoothed Approach to Multi-view Subspace Clustering](https://mlsanity.com/paper/2203.08060/) (2022)
- [An Image Clustering Auto-Encoder Based on Predefined Evenly-Distributed Class Centroids and MMD Distance](https://mlsanity.com/paper/1906.03905/) (2019)
- [Unsupervised Feature Learning Architecture with Multi-clustering Integration RBM](https://mlsanity.com/paper/1812.01967/) (2018)

## Built on by (most cited first, 5 of 5)

- [Fuzzy Cluster-Aware Contrastive Clustering for Time Series](https://mlsanity.com/paper/2503.22211/) (2025)
- [Deep Representation Learning-Based Dynamic Trajectory Phenotyping for Acute Respiratory Failure in Medical Intensive Care Units](https://mlsanity.com/paper/2405.02563/) (2024)
- [Concrete Dense Network for Long-Sequence Time Series Clustering](https://mlsanity.com/paper/2405.05015/) (2024)
- [From Patterns to Predictions: A Shapelet-Based Framework for Directional Forecasting in Noisy Financial Markets](https://mlsanity.com/paper/2509.15040/) (2025)
- [AttentiveGRUAE: An Attention-Based GRU Autoencoder for Temporal Clustering and Behavioral Characterization of Depression from Wearable Data](https://mlsanity.com/paper/2510.02558/) (2025)

## Abstract (arXiv)

Recently the deep learning has shown its advantage in representation learning and clustering for time series data. Despite the considerable progress, the existing deep time series clustering approaches mostly seek to train the deep neural network by some instance reconstruction based or cluster distribution based objective, which, however, lack the ability to exploit the sample-wise (or augmentation-wise) contrastive information or even the higher-level (e.g., cluster-level) contrastiveness for learning discriminative and clustering-friendly representations. In light of this, this paper presents a deep temporal contrastive clustering (DTCC) approach, which for the first time, to our knowledge, incorporates the contrastive learning paradigm into the deep time series clustering research. Specifically, with two parallel views generated from the original time series and their augmentations, we utilize two identical auto-encoders to learn the corresponding representations, and in the meantime perform the cluster distribution learning by incorporating a k-means objective. Further, two levels of contrastive learning are simultaneously enforced to capture the instance-level and cluster-level contrastive information, respectively. With the reconstruction loss of the auto-encoder, the cluster distribution loss, and the two levels of contrastive losses jointly optimized, the network architecture is trained in a self-supervised manner and the clustering result can thereby be obtained. Experiments on a variety of time series datasets demonstrate the superiority of our DTCC approach over the state-of-the-art.