1. Dataset

2. Workflow

Extract suspected samples from data using k-NN,
test AutoEncoder against anomalous data, assuming that k-NN's decision has been correct,
compare results of three models: k-NN, AutoEncoder and ensemble IsolationForest

Name		Name	Last commit message	Last commit date
Latest commit History 10 Commits
.gitignore		.gitignore
Autoencoder Parameter Search.xlsx		Autoencoder Parameter Search.xlsx
LICENSE		LICENSE
Network Anomaly Detection.pdf		Network Anomaly Detection.pdf
README.md		README.md
[1] Data Preprocessing.ipynb		[1] Data Preprocessing.ipynb
[2] Unsupervised k-NN.ipynb		[2] Unsupervised k-NN.ipynb
[3] Deep AutoEncoder.ipynb		[3] Deep AutoEncoder.ipynb
[4] Isolation Forest and Final Comparison.ipynb		[4] Isolation Forest and Final Comparison.ipynb
deep_autoencoder.py		deep_autoencoder.py