iotoutlier1 0.0.1

Data representation for IoT traffic

Purpose

"Comparison of data representation for outlier detection in IOT"  

1. Requirements (pip3 freeze > requirements.txt)

python==3.7.4
seaborn==0.9.0
scipy==1.3.1
matplotlib==3.1.1
scapy==2.4.3
pandas==0.25.1
pyod==0.7.4
numpy==1.17.0
umap_learn==0.3.10
scikit_learn==0.21.3
umap==0.1.1
keras==2.3.1  # for autoencoder
tensorflow==2.0.0

2. Project Directory Structure

iot_outlier_src: source root directory (set 'iot_outlier_src' as sources root)

|- input_data: raw data

if any file is more than 100MB, please do not store it at here

|- output_data: results

figures: store roc and auc
model_dumpling: store models to disk
...

|- data_process:

    pcap2features
    features2dataset

|- detector

gmm
ocsvm

|- utils

toolkit to preprocess input data, such as 'load data', 'normalization data'
visualization: plot data to visualize 
    ..

|- log: use to log middle or tmp results.

...

|-legacy: backup

...

Main

main_featcomp-class.py

docs

To do ...

tests

To do ... 

Note: