DISCONTINUATION OF PROJECT.

This project will no longer be maintained by Intel.

Intel has ceased development and contributions including, but not limited to, maintenance, bug fixes, new releases, or updates, to this project.

Intel no longer accepts patches to this project.

If you have an ongoing need to use this project, are interested in independently developing it, or would like to maintain patches for the open source software community, please create your own fork of this project.

Link to MLsploit Main Project

MLsploit Module: Resilient-ML-Research-Platform

This is a web platform to demo Machine Learning as a Service (MLaaS) on security researches. It has a machine learning (ML) pipeline to build and tune models. It also has a portal to demo adversarial ML and countermeasures.

• MLaaS:
  • ML classifier creation and inference
• MLsploit:
  • Adversarial ML and demos

Getting Started

Dependancies

• CentOS or Ubuntu
• Apache Hadoop, Spark & MongoDB
• Python 2.7 and related Python packages, e.g. Scikit-Learn, Numpy, Keras etc
• Django & SQLite
• Docker (optional for demo containers)

Installation

• Demo cluster by Docker containers - tiny bigdata platform on your Linux laptop:

docker login                    # Login to Docker Hub by your id & password
cd ./docker                     # cd to folder "docker" in git cloned project
chmod 755 *.sh                  # Change scripts to be executable
sudo ./setup_docker_linux.sh    # Create users on Linux and copy related files
./run_container_linux.sh        # Pull images from Docker Hub and run 4 containers:
                                #   HDFS/Spark master & slave1, mongo & Django web 
                                # Access at http://<your machine dns>:8000/ id=demo pwd=demo123

• For full installation, please follow the Setup_Guide_CentOS7.pdf
  • Modify web configuation files for setting Hadoop/Spark/web/MongoDB hostnames
    • app.config
    • myml/settings.py
    • atdml/settings.py etc.

Design Diagrams

Data Flow:

Software Stack:

Note: DNN worker to be released...

License

This project is licensed under the Apache 2.0