Recent Employment

 
Dartmouth.png

Dartmouth College

Security of Machine Learning Systems | cURRENT

Working with George Cybenko at exploring the security of machine learned systems through methods connected to formal languages and algebraic topology.

 
str.png

Systems & Technology Research

Integer Discrete Flows for Sequential Data | June 2020 - August 2020, December 2020 - January 2021 

Introduced and studied a variant of Hoogebloom et al.’s integer discrete flows (IDFs) to work on variable length sequential data.

 
DoD.png

Department of Defense

Multi-Modal Similarity Search | December 2019 - January 2020, June 2019 - August 2019

Applied my multidisciplinary experiences in collaboration with defense contractors and government customers towards developing next generational capabilities to support Department of Defense missions.

Recent Projects

 
Jammer.png

 Poisoning Q-Learned Anti-Jamming Systems

Poisoning Attacks on Reinforcement Learned Jamming Defenses | September 2020 - November 2020

Demonstrated the effectiveness of physically realizable adversarial machine learning attacks towards disarming Q-learning anti-jamming defenses. Roughly speaking, our attacks ‘teaches’ the defending system to avoid sending packets by poisoning the environmental rewards. As such, the jammer can achieve their goal with less energy, and less engagement.

 
Virus.png

Tanzania During COVID-19

Superstition, Disinformation, and Political Opportunity | March 2020 - May 2020

Web scraped gigabytes of data off the Tanzanian social network Jamii Forms, and performed an exploratory data analysis to understand the political landscape of Tanzania during the COVID-19 global pandemic. Used a wide range of analytic tools from multi-lingual NLP to graph metrics like eigen-centrality, and found significant disapproval of John Magufuli, and negativity around COVID-19.

 
Graph.png

ANALYSIS OF New-Type Warfare Attacks on America’s Agricultural SYSTEM

Cybersecurity and information warfare in America’s Agricultural Sector

Applied my technical experience with disruptive technologies in a policy context to study how precision agricultural technologies are making America’s agricultural sector vulnerable to asymmetric warfare tactics. Proposed concrete, technical research directions which will be necessary to increase our resiliency to such tactics.

Education

Dartmouth COLLEGE

PhD in Computer Science
Current Student

Carnegie Mellon UNIVERSITY

Discrete Math & Logic with Minor in Computer Science
Graduated May 2015

Talks

woRK TO BE SHOWN AT iNVITED TALK AT lANGSEC 2021

WORK TO BE SHOWN AT SIAM international conference on data mining

Contact

Email: joshua.m.ackerman.gr@dartmouth.edu