ECE 579M ST: Machine Learning in Cybersecurity

When:          Tuesdays 6:00 – 8:50 pm

Where:        AK 219

Instructor:  Koksal Mus kmus@wpi.edu  / AK 307

T.A. :  Berk Gulmezoglu bgulmezoglu@wpi.edu   / AK 212B

Office Hour: Monday 4:00-5:00pm

Syllabus

Jan 16 in class reading materials

  • Nielsen's Online Book Chapter about recognizing handwritten digits

    Jan 23 in class reading materials

  • A Step by Step Backpropagation Example

  • Explaining And Harnessing Adversarial Examples

  • Practical Black-Box Attacks against Machine Learning

    Jan 30 in class reading materials

  • One pixel attack for fooling deep neural networks

  • Delving Into Transferable Adversarial Examples And Black-Box Attacks

    Feb 6 in class reading materials

  • Distilling the Knowledge in a Neural Network

  • Distillation as a Defense to Adversarial Perturbations against Deep Neural Networks

    Feb 13 in class reading materials

  • Defensive Distillation is Not Robust to Adversarial Examples

  • Machine Learning Models that Remember Too Much

    Feb 20 in class reading materials

  • Evading Classifiers by Morphing in the Dark

  • Cache-Based Application Detection in the Cloud Using Machine Learning

  • PerfWeb: How to Violate Web Privacy with Hardware Performance Events

    Feb 27 in class reading materials

  • Malware Detection using Machine Learning Based Analysis of Virtual Memory Access Patterns

  • Evading Machine Learning Malware Detection

  • Adversarial Examples for Malware Detection

    March 20 in class reading materials

  • Progressive Growing of GANs for Improved Quality, Stability, and Variation

  • Generative Adversarial Nets

  • PassGAN: A Deep Learning Approach for Password Guessing

    March 27 in class reading materials

  • Beauty and the Burst: Remote Identification of Encrypted Video Streams

  • Fake News Detection on Social Media:A Data Mining Perspective

    April 03 in class reading materials

  • Predicting Domain Generation Algorithms with Long Short-Term Memory Networks

  • An Analysis of Recurrent Neural Networks for Botnet Detection Behavior

  • Application of Recurrent Neural Networks for User Verification based on Keystroke Dynamics

    Presentations (April 17)

    Membership Inference Attacks Against Machine Learning Models

    Detection of Distributed Denial of Service Attacks Using Artificial Neural Networks

    Presentations (April 24)

    Employing Machine Learning Techniques for Detection and Classification of Phishing Emails

    Adversarial Learning as Defense to Side-channel Attacks

    Logistic regression over encrypted genomic data

    Projects

    Project 1 (Jan 30)

  • Building MNIST Classifier

    look at MNIST Experiments of Nielsen Chapter one ( Implementing our network to classify digits )

    Project 2 (Feb 13)

  • Adversarial Crafting I(Submit it to Canvas)

    Project 3 (Feb 27)

  • Machine Learning on Side-Channel Attacks (Submit it to Canvas)

    Project 4 (April 3)

  • Privacy Violation Analysis By Using Regression (Submit it to Canvas)

    Project 5 (April 24)

  • Machine Learning on Intrusion Detection (Submit it to Canvas)

    Reading Materials

    1. Nielsen’s Online Book

    2. Goodfellow’s Online book

    3. DeepFool: a simple and accurate method to fool deep neural networks

    4. Evading Classifiers by Morphing in the Dark

    5. Machine Learning Models that Remember Too Much

    6. Deep Models Under the GAN: Information Leakage from Collaborative Deep Learning

    7. One pixel attack for fooling deep neural networks

    8. Oblivious Neural Network Predictions via MiniONN Transformations

    9. Query-limited Black-box Attacks to Classifiers

    10. Towards evaluating the robustness of neural networks

    11. MagNet: a Two-Pronged Defense against Adversarial Examples

    12. DolphinAtack: Inaudible Voice Commands

    13. Note on Attacking Object Detectors with Adversarial Stickers

    14. Adversarial Examples: Attacks and Defenses for Deep Learning

    15. When Not to Classify: Anomaly Detection of Attacks (ADA) on DNN Classifiers at Test Time

    16. Introduction to Adversarial Machine Learning

    17. Robust Adversarial Examples

    18. Attacking Machine Learning with Adversarial Examples

    19. Breaking Linear Classifiers on ImageNet

    20. Intriguing properties of neural networks

    21. Explaining And Harnessing Adversarial Examples

    22. The Limitations of Deep Learning in Adversarial Settings

    23. Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samples

    24. Delving Into Transferable Adversarial Examples And Black-Box Attacks

    25. Practical Black-Box Attacks against Machine Learning

    26. Adversarial Machine Learning At Scale

    27. Ensemble Adversarial Training: Attacks and Defenses

    28. Classifiers Under Attack

    29. Adversarial Examples In Machine Learning

    30. Adversarial Learning for Good: My Talk at #34c3 on Deep Learning Blindspots