# Harnessing Contrastive Learning and Neural Transformation for Time Series Anomaly Detection

arXiv: https://arxiv.org/abs/2304.07898 · HTML: https://mlsanity.com/paper/2304.07898/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 22 of its 36 references that are in the map.
- First posted on arXiv 16 Apr 2023.

## Builds on (most cited first, 22 of 22)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [Representation Learning with Contrastive Predictive Coding](https://mlsanity.com/paper/1807.03748/) (2018)
- [Momentum Contrast for Unsupervised Visual Representation Learning](https://mlsanity.com/paper/1911.05722/) (2019)
- [VICReg: Variance-Invariance-Covariance Regularization for Self-Supervised Learning](https://mlsanity.com/paper/2105.04906/) (2021)
- [Using Self-Supervised Learning Can Improve Model Robustness and Uncertainty](https://mlsanity.com/paper/1906.12340/) (2019)
- [Detecting Spacecraft Anomalies Using LSTMs and Nonparametric Dynamic Thresholding](https://mlsanity.com/paper/1802.04431/) (2018)
- [Anomaly Transformer: Time Series Anomaly Detection with Association Discrepancy](https://mlsanity.com/paper/2110.02642/) (2021)
- [Graph Neural Network-Based Anomaly Detection in Multivariate Time Series](https://mlsanity.com/paper/2106.06947/) (2021)
- [Deep Anomaly Detection Using Geometric Transformations](https://mlsanity.com/paper/1805.10917/) (2018)
- [LSTM-based Encoder-Decoder for Multi-sensor Anomaly Detection](https://mlsanity.com/paper/1607.00148/) (2016)
- [TranAD: Deep Transformer Networks for Anomaly Detection in Multivariate Time Series Data](https://mlsanity.com/paper/2201.07284/) (2022)
- [A Deep Neural Network for Unsupervised Anomaly Detection and Diagnosis in Multivariate Time Series Data](https://mlsanity.com/paper/1811.08055/) (2018)
- [MAD-GAN: Multivariate Anomaly Detection for Time Series Data with Generative Adversarial Networks](https://mlsanity.com/paper/1901.04997/) (2019)
- [Contrastive Learning of General-Purpose Audio Representations](https://mlsanity.com/paper/2010.10915/) (2020)
- [Learning Graph Structures with Transformer for Multivariate Time Series Anomaly Detection in IoT](https://mlsanity.com/paper/2104.03466/) (2021)
- [Neural Transformation Learning for Deep Anomaly Detection Beyond Images](https://mlsanity.com/paper/2103.16440/) (2021)
- [An Evaluation of Anomaly Detection and Diagnosis in Multivariate Time Series](https://mlsanity.com/paper/2109.11428/) (2021)
- [Deep Learning for IoT Big Data and Streaming Analytics: A Survey](https://mlsanity.com/paper/1712.04301/) (2017)
- [Deep Contrastive One-Class Time Series Anomaly Detection](https://mlsanity.com/paper/2207.01472/) (2022)
- [Multivariate Time Series Anomaly Detection via Dynamic Graph Forecasting](https://mlsanity.com/paper/2302.02051/) (2023)
- [DeepChannel: Salience Estimation by Contrastive Learning for Extractive Document Summarization](https://mlsanity.com/paper/1811.02394/) (2018)

## Built on by (most cited first, 1 of 1)

- [TFEC: Multivariate Time-Series Clustering via Temporal-Frequency Enhanced Contrastive Learning](https://mlsanity.com/paper/2601.07550/) (2026)

## Abstract (arXiv)

Time series anomaly detection (TSAD) plays a vital role in many industrial applications. While contrastive learning has gained momentum in the time series domain for its prowess in extracting meaningful representations from unlabeled data, its straightforward application to anomaly detection is not without hurdles. Firstly, contrastive learning typically requires negative sampling to avoid the representation collapse issue, where the encoder converges to a constant solution. However, drawing from the same dataset for dissimilar samples is ill-suited for TSAD as most samples are ``normal'' in the training dataset. Secondly, conventional contrastive learning focuses on instance discrimination, which may overlook anomalies that are detectable when compared to their temporal context. In this study, we propose a novel approach, CNT, that incorporates a window-based contrastive learning strategy fortified with learnable transformations. This dual configuration focuses on capturing temporal anomalies in local regions while simultaneously mitigating the representation collapse issue. Our theoretical analysis validates the effectiveness of CNT in circumventing constant encoder solutions. Through extensive experiments on diverse real-world industrial datasets, we show the superiority of our framework by outperforming various baselines and model variants.