Deep Learning Platform for network analysis This is an ongoing project. The initial goal is to build a deep learning platform and network ddos traffic analysis system.

Building a Machine Learning based Off-line Network Analysis and Attack Detection Application.

Training and Testing Model Management and Storing.

Data Preprocessing, Manipulating and Visualization.

Full Stack Web Service for Deep Learning.

The architecture of our system includes a front end framework (Angular.js), a back end server (Node.js and express), MongoDB Database, Data Analysis Engine, Defence Engine. For the web framework, we use the mean stack framework (mongoDB, Express, Angular.js and Node.js).

Front end GUI: The function of the front end is to show a dashboard of the current training and testing result of the analysis.

Back End Server: Back-end server takes the responsibility to get response to the front-end website, get access to the database and run the data statistic engine.

Deep Learning Engine: The deep learning enigine will make use of the Hadoop clusters to train the model in a more efficient way.