PyOD: A Python Toolbox for Scalable Outlier Detection

By Yue Zhao, Zain Nasrullah, Zheng Li. 2019

903 citations recorded by the source corpus.

PyOD is an open-source Python toolbox for performing scalable outlier detection on multivariate data. Uniquely, it provides access to a wide range of outlier detection algorithms, including established outlier ensembles and more recent neural network-based approaches, under a single, well-documented API designed for use by both practitioners and researchers. With robustness and scalability in mind, best practices such as unit testing, continuous integration, code coverage, maintainability checks, interactive examples and parallelization are emphasized as core components in the toolbox's development. PyOD is compatible with both Python 2 and 3 and can be installed through Python Package Index (PyPI) or this https URL.

Subfields: Artificial Intelligence, Public Health, Environmental and Occupational Health, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Sanjay Krishnan
2. Eugene Wu
3. Xiyang Hu
4. Songqiao Han
5. Hailiang Huang
6. Mingqi Jiang
7. Yue Zhao
8. Guansong Pang

In Public Health, Environmental and Occupational Health

1. Eugene Wu
2. Sanjay Krishnan
3. Xiyang Hu
4. Yue Zhao

In Signal Processing

1. Eugene Wu
2. Sanjay Krishnan
3. Xiyang Hu
4. Yue Zhao
5. Bin Cui
6. Ming-Hsuan Yang
7. Yongjun Wang
8. Chuan-Sheng Foo

Citations in the corpus

• AlphaClean: Automatic Generation of Data Cleaning Pipelines (2019, methodology): “The default library consists of parametrized outlier detector methods from dBoost [34] and pyod [49].”
• ADBench: Anomaly Detection Benchmark (2022, methodology): “Some existing works, such as PyOD [198], generate fully synthetic anomalies by assuming their data distribution, which fails to create complex anomalies.”
• COPOD: Copula-Based Outlier Detection (2020, methodology): “Outlier detection algorithms have many practical applications [1], such as credit card fraud detection, network intrusion detection, and synthetic data generation [2].”
• Learning to Classify With Incremental New Class (2021, neutral): “D ISCUSSION Novelty detection, a.k.a. anomaly detection [18], [9], [57]–[62] refers to the process of detecting data instances that significantly deviate from the majority of data instances.”
• Diffusion Models: A Comprehensive Survey of Methods and Applications (2022, background): “Anomaly detection is a critical and challenging problem in machine learning [209, 283] and computer vision [262].”
• Progress in Outlier Detection Techniques: A Survey (2019, neutral): “2) Python Outlier Detection (PyOD) [232]”
• Deep Isolation Forest for Anomaly Detection (2022, background): “Also, some practical anomaly detection tools, e.g., [19], [20], [21], [22], are developed to facilitate the use of anomaly detection models in real-world applications.”
• ECOD: Unsupervised Outlier Detection Using Empirical Cumulative Distribution Functions (2022, background): “Notably, PyOD is a popular open-source Python toolbox for performing scalable outlier detection on multivariate data [60].”
• An Evaluation of Anomaly Detection and Diagnosis in Multivariate Time Series (2021, background): “CPS unlike many distance-based and pattern-based techniques that are not applicable to streaming time-series, as they require both training and test data during inference [24]–[27], and (c) the ability to localize anomalous time points within sequences, unlike techniques [24], [25], [31] that work a”
• Do You Know What You Are Talking About? Characterizing Query-Knowledge Relevance For Reliable Retrieval Augmented Generation (2024, methodology): “We use the PyOD package (Zhao et al., 2019) with default parameters for implementing these outlier detection baselines.”