Robustness and Security in ML Systems: Junfeng Yang

E6998 Robustness and Security in ML Systems

Spring 2018 -- Junfeng Yang

We put a tentative syllabus here to give you an idea what future may bring. This syllabus is subject to change as the course progresses.


# Day Date Topic Assignment Speakers
1 Mon Jan 22 Introduction Form reading group
2 Mon Jan 29 Deep learning Read Lecun-90c, AlexNet Ben, Ted
3 Mon Feb 5 Adversarial ML (1) Read Adversarial ML, Intriguing properties of neural networks Kai-zhan, Sid
4 Mon Feb 12 Adversarial ML (2) Read Fast adversarial DL, Blackbox adversarial DL Dongdong
5 Mon Feb 19 Interpreting DL (1) Read Extracting NN rules, NN visual representations Mohamed, Michael
6 Mon Feb 26 Interpreting DL (2) Read Influence functions, NN interpretations fragile Bryan, Eduardo
7 Mon Mar 5 Verifying DL (1) Read NN safety verification, Reluplex Shiqi, Maryam
8 Mon Mar 12 No class (Spring recess)
9 Mon Mar 19 Verifying DL (2) Read DeepSafe, Measuring NN robustness Himanshu, Linjie
10 Mon Mar 26 DL testing and abstract intepretation Read Featured-guided black-box safety testing, AI2: Abstract Interpretation of Neural Networks Raphael, Daniel
11 Mon Apr 2 Robustness training Read Robustness optimization, Stability training Bryan, Raphael; Eduardo, Mohamed
12 Mon Apr 9 Robustness by construction Read Provable defenses, Certifiable robustness Daniel, Shiqi; Maryam, Michael
13 Mon Apr 16 Neural programmer-interpreters Read Neural programmer-interpreters, NPI with recursion Ted, Himanshu; Linjie, Siddharth
14 Mon Apr 23 Testing DL Read DeepXplore, VeriVis
15 Mon Apr 30 Mini-research conference Present and demo your final project