Anomaly Detection
Algorithms in Machine Learning, ISAE-SUPAERO
This repository contains course materials for my Machine Learning class on the topic of Anomaly Detection:
- anomaly_detection_class.pdf : course presentation with teaching content
- anomaly_detection_class.ipynb : Python notebook illustrating the course notions
- anomaly_detection_class-empty.ipynb : same Python notebook to follow with students during class
An application exercise is also available:
- anomaly_detection_exercise.pdf : instructions for the exercise
- data/dataset.csv : the dataset to be used to answer the questions
Option 1: Working with Google Colab
To follow the notebooks with Google Colab, simply go to https://colab.research.google.com/. Import a new notebook from GitHub, search for "jfabrice" and open one of the notebooks of this class (ml-class-anomaly-detection), for example anomaly_detection_class-empty.ipynb. Then click on "Copy to Drive" to be able to execute it. The first section of the notebook is there to initialize the environment from Google Colab.
Option 2: Working locally - Setting up Anaconda environment
To setup the Anaconda environment with required dependencies, execute the following instructions in Anaconda prompt or Linux shell.
# Clone this github repository on your machine
git clone https://github.com/jfabrice/ml-class-anomaly-detection.git
# Change working directory inside the repo
cd ml-class-anomaly-detection
# Create a new virtual environment
conda create -n anomalydetectionenv python==3.6
# Activate the environment
## For Linux / MAC
source activate anomalydetectionenv
## For Windows
activate anomalydetectionenv
# Install the requirements
pip install -r requirements.txt