Skip to content

README.md

Latest commit

 

History

History
45 lines (33 loc) · 1.18 KB

README.md

File metadata and controls

45 lines (33 loc) · 1.18 KB

Network Traffic Classification and Anomaly Detection

Deep Learning model for analysis and identify the application for given Teletraffic pattern. Experiment with different models including MLPs and CNNs. Final object is to detect anomaly apps with unusual traffic patterns.

  • Analyze network traffic for both incoming and outgoing
  • Extract statistical features
  • Train the supervised deep learning model
  • Handle anomalies using softmax probabilities

Techniques

  • Supervised Deep Learning, Unsupervised Deep Learning
  • Statistical feature calculation

Models

  • 1D-CNN with Statistical features
  • ANN with Statistical features
  • 2D-CNN with Spatial features

Tools

  • TensorFlow - Deep Learning Model
  • pandas - Data Extraction and Preprocessing
  • numpy - numerical computations
  • scikit learn - Advanced preprocessing

Installation

Install the dependencies and conda environment

$ conda create -n envname python=python_version
$ activate envname 
$ conda install -c anaconda tensorflow-gpu
$ conda install -c anaconda pandas
$ conda install -c anaconda matplotlib
$ conda install -c anaconda scikit-learn

For Train Model...

$ python model.py