/anomaly-detection-code-pattern

Sample Jupyter Notebook for playing around with the Anomaly Detection service to be made available on API Hub

Primary LanguageJupyter NotebookApache License 2.0Apache-2.0

anomaly-detection-code-pattern

Sample Jupyter Notebooks for playing around with the IBM Anomaly Detection service to be made available on IBM API Hub.

Requirements and Installation

  1. Python 3
  2. Credentials to access the API service (Please follow the instructions)
  3. Clone the repository
git clone https://github.com/IBM/anomaly-detection-code-pattern.git
cd anomaly-detection-code-pattern/
  1. (Optional) Create a virtual environment
virtualenv ad_env
source ad_env/bin/activate
  1. Install required packages
pip install -r requirements.txt
  1. Open Jupyter notebook in current directory
python -m ipykernel install --user --name=ad_env  # optional: add virtual environment to jupyter notebook
jupyter notebook

Notebooks

Here are the list of provided notebooks:

  1. Univariate_AD_service_public_data.ipynb: Anomaly detection on univariate public data
  2. Univariate_AD_service_sample_data.ipynb: Anomaly detection on univariate sample data
  3. Multivariate_AD_service_sample_data.ipynb: Anomaly detection on multivariate sample data
  4. Regression-aware_AD_service_sample_data.ipynb
  5. MixtureModel-aware_AD_service_sample_data.ipynb

Additional Links

  1. API Service in IBM API Hub
  2. API Service in IBM Learning Path

Talks

  1. MLSys 2022 half day tutorial
  2. KDD 2022 3-hour tutorial
  3. ICDE 2022 tutorial
  4. DASFAA 2022 tutorial