11488/11688/11788 — Spring 2020

Computational Forensics and AI

With new AI-based technologies that power almost all activities in the digital world, cybercrime is on an unprecedented increase. Forensics is the science of tracing causes, methods and perpetrators from evidence, once a crime has been committed. This course will teach you some of the key technologies that are being used to track cybercriminals.

Instructor and TAs

Instructor: Rita Singh (rsingh@cs.cmu.edu)

Shahan Ali Memon (samemon@cs.cmu.edu)
Mahmoud Al-Ismail (mahmoudi@andrew.cmu.edu)
Bhiksha Raj (bhiksha@cs.cmu.edu)
Ben Striner (bstriner@andrew.cmu.edu)

Venue and timings for lectures and office hours

Lecture: Monday and Wednesday, 9:00 a.m. - 10:20 a.m. @WEH 5409
Office Hours: By appointment

Basic course structure

There will be 14 weeks of lectures. A different forensic subarea will be covered each week in 2 lectures of 1 hr 20 mins each. There will be one quiz at the end of each week (released Friday midnight, due Sunday midnight) and 4 homeworks in all. Each homework will be due within 2 weeks from the date of release.

Schedule of lectures:

Week 1: Introduction
Week 2: Network forensics
Week 3: Dark Web forensics (HW1 released)
Week 4: Machine learning in forensics
Week 5: Deep learning and AI in forensics
Week 6: Cryptography (HW2 released)
Week 7: Image forensics
Week 8: Steganography
Week 9: Text and social media forensics (HW3 released)
Week 10: Video forensics
Week 11: Audio forensics
Week 12: DeepFakes: generation and tracking (HW4 released)
Week 13: Computer forensics
Week 14: Advanced tracking techniques

A more comprehensive syllabus can be found here

The weekly lectures will be slide-based. Research papers, (accessible) textbook chapters and notes will be provided as links where necessary. Slides for each class will be uploaded on Piazza after each class, on the same day as the class. Homeworks will be released in class, and explained in a 30 minute recitation at the end of the corresponding class. Attendance is expected and recommended. Lectures will not be recorded.

Enrollment requirements

You must know programming (preferably Python). Basic skills in maths, statistics and probability are expected.