Building a large scale unsupervised model anomaly detection system

Welcome to the Anomaly Detection Demo!

Prerequisites: Docker Desktop or Docker daemon running on your machine.

Using the pre-built docker container

We have provided a docker image that you can run on your local machine and follow along the notebooks with the instructor.

The data files needed are also loaded into the docker image.

export DOCKER_DEFAULT_PLATFORM=linux/amd64

export VERSION=2.4

docker run -it --rm -p 8888:8888 anindyas/pydata-seattle:"${VERSION}"

Using --env GRANT_SUDO=yes --user root will spawn the jupyter notebook with jovyan having root privileges.

Due to the usage of the flag --rm Docker automatically cleans up the container and removes the file system when the container exits.