# Unsupervised Feature Construction for Anomaly Detection in Time Series -- An Evaluation

arXiv: https://arxiv.org/abs/2501.07999 · HTML: https://mlsanity.com/paper/2501.07999/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 34 references that are in the map.
- First posted on arXiv 14 Jan 2025.

## Builds on (most cited first, 12 of 12)

- [ADBench: Anomaly Detection Benchmark](https://mlsanity.com/paper/2206.09426/) (2022)
- [A review on outlier/anomaly detection in time series data](https://mlsanity.com/paper/2002.04236/) (2020)
- [Time-Series Anomaly Detection Service at Microsoft](https://mlsanity.com/paper/1906.03821/) (2019)
- [Current Time Series Anomaly Detection Benchmarks are Flawed and are Creating the Illusion of Progress](https://mlsanity.com/paper/2009.13807/) (2020)
- [catch22: CAnonical Time-series CHaracteristics](https://mlsanity.com/paper/1901.10200/) (2019)
- [Towards a Rigorous Evaluation of Time-series Anomaly Detection](https://mlsanity.com/paper/2109.05257/) (2021)
- [Anomaly Detection in Univariate Time-series: A Survey on the State-of-the-Art](https://mlsanity.com/paper/2004.00433/) (2020)
- [TadGAN: Time Series Anomaly Detection Using Generative Adversarial Networks](https://mlsanity.com/paper/2009.07769/) (2020)
- [Dive into Time-Series Anomaly Detection: A Decade Review](https://mlsanity.com/paper/2412.20512/) (2024)
- [The FreshPRINCE: A Simple Transformation Based Pipeline Time Series Classifier](https://mlsanity.com/paper/2201.12048/) (2022)
- [Open challenges for Machine Learning based Early Decision-Making research](https://mlsanity.com/paper/2204.13111/) (2022)
- [Automatic Feature Engineering for Time Series Classification: Evaluation and Discussion](https://mlsanity.com/paper/2308.01071/) (2023)

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

- [Evaluating the Sensitivity of BiLSTM Forecasting Models to Sequence Length and Input Noise](https://mlsanity.com/paper/2512.06926/) (2025)

## Abstract (arXiv)

To detect anomalies with precision and without prior knowledge in time series, is it better to build a detector from the initial temporal representation, or to compute a new (tabular) representation using an existing automatic variable construction library? In this article, we address this question by conducting an in-depth experimental study for two popular detectors (Isolation Forest and Local Outlier Factor). The obtained results, for 5 different datasets, show that the new representation, computed using the tsfresh library, allows Isolation Forest to significantly improve its performance.