Anyscale Learning For All

Machine learning technology, evolutionary algorithms, and data science for knowledge mining, prediction, analytics, and optimization with projects in cyber security, software analysis, MOOC technology, and medical technology.

 

Research Themes

Artificial Adversarial Intelligence

Understanding Programs & Programming

Data Analytics
Artificial Adversarial Intelligence Understanding Programs and Programming Data Analytics

 

Applications: Adversarial dynamics modeling of networks, adversarial examples in malware and software, adversarial hardening, malware detection and classification, automatic program analysis for vulnerabilities, security by deception, security by segmentation, and peer to peer network defenses against DDOS, lateral stage and scanning attacks; MOOCs on computational thinking and programming, MOOC clickstream and activity data analysis with deep learning; medical experimental analytics, COVID-19 agent-based modeling.

 

 

Featured ALFA News!

ALFA welcomes Stepken Moskal, Postdoc Associate

Stephen Moskal (He/Him) is a Ph.D. of Engineering from the Rochester Institute of Technology and has joined the MIT CSAIL's ALFA group at the beginning of 2022. Stephen's background is in Computer Engineering and Machine Learning and he specializes in the understanding, modelling, and simulation of cyber attack behaviors. Stephen has experience in industry applications of cyber-research and start-up incubator experience combined with his academic experience. It is due to this that Stephen focuses on creating solutions to difficult cyber security problems that can be realistically be deployed given the often limited resources of security analysts.

Some of his research includes:
CASCADES (Cyber Attack Scenario and Network Defense Simulator) - A robust cyber attack behavior simulator to find possible attack paths on a virtual network.
Moskal, Stephen, Shanchieh Jay Yang, and Michael E. Kuhl. "Cyber threat assessment via attack scenario simulation using an integrated adversary and network modeling approach." The Journal of Defense Modeling and Simulation 15.1 (2018): 13-29.

PATRL (Psuedo-Active Transfer Learning) - Interprets descriptions of cyber attacks into cyber attack kill chain stages. Uses unsupervised deep-language modelling of cyber-contextual texts and "Psuedo-Active Learning" to significantly reduce the amount of labelled data to be provided by an analyst.
Moskal, Stephen, and Shanchieh Jay Yang. "Translating Intrusion Alerts to Cyberattack Stages using Pseudo-Active Transfer Learning (PATRL)." 2021 IEEE Conference on Communications and Network Security (CNS). IEEE, 2021.

HeAT (Heated Alert Triage) - Uses prior triages of intrusion alerts combined with network-agnostic features to quickly identify future attack campaigns. Our process to extract out the analyst's intuition with the network-agnostic features enabled HeAT to correctly identify attack campaigns regardless of the network.
Moskal, Stephen Frank. HeAt PATRL: Network-Agnostic Cyber Attack Campaign Triage with Pseudo-Active Transfer Learning. Diss. Rochester Institute of Technology, 2021.

Linkedin: https://www.linkedin.com/in/stephen-moskal/ 
Music: https://www.mixcloud.com/shadw_moses/ 

 

 

 

Diversity Drives Innovation and Learning

At the Anyscale Learning For All (ALFA) Group, we mean Anyscale Learning for ALL. ALFA is dedicated to cultivating an inclusive culture that supports, promotes, and empowers diverse voices in Computer Science & AI. Our focus is to improve and build machine learning, AI, and data analytics technology that works for everyone.

Representation matters.

We want our research to be representative of everyone who benefits and learns from it. We value people with different experiences, perspectives, and backgrounds - it’s the cornerstone of our approach to learning and research. We celebrate diversity along many axes: race, religion, ethnicity, age, sex, national origin, sexual orientation, gender identity, gender expression, genetic disposition, neurodiversity, disability, veteran status and any other aspect which makes you unique.

 

 

ALFA's research would not be possible without the past and ongoing support of our industry sponsors
ALFA sponsors past and present
The views, opinions and positions expressed by ALFA Group and on this site are theirs alone, and do not necessarily reflect the views, opinions or positions of their sponsors
 
 
 
ALFA Group is formerly the Evolutionary Design and Optimization (Evo-DesignOpt) Group
 
 

 

Recent News

2022

April

Seeking Deep Understanding in Machine Learning
Podcast Spotlight
Una-May O'Reilly
MIT-CSAIL Alliances

January

ALFA welcomes a new postdoc, Stephen Moskal.
Stephen is a graduate of Rochester Institute of Technology. His research career throughout his MS and PhD degrees involved the understanding of cyber attack behaviors, how to identify them, and then simulate their impacts. His dissertation (defended in Dec. 2021) called “HeAT-PATRL” leveraged multiple deep learning AI techniques to discover cyber attack campaigns with in real-world IDS (Intrusion Detection Systems) dataset. His work focuses on practical and deployable solutions to cyber attack problems.

 

2021

October

Adversarial Intelligence
Una-May O'Reilly
Artificial Intelligence in the Liberal Arts Series, Amhurst College, 2021

 

September

ALFA welcomes new PhD Candidates Aruna Sankaranarayanan & Michael Wang

 

Artificial Adversarial Intelligence
Una-May O'Reilly
Keynote, UK Workshop on Computation Intelligence, UKCI, 2021. 

 

July

ALFA PhD Candidate, Shashank Srikant's research in collaboration with the MIT-IBM Watson AI Lab is featured in The Economist and WIRED!

 
 AI is transforming the coding of computer programs (full article)

 
 Now for AI’s Latest Trick: Writing Computer Code (full article)

 

June

ALFA welcomes summer UROPs: Ness Vera, Muhua Xu, Eileen Li. We bid farewell to research intern, Sasha Shashkov.

Congratulations to Diana Flores for the completion of her M.Eng degree. 

 

Artificial Adversarial Intelligence: Code Model Attacks and Robustness
Una-May O'Reilly
Workshop on Adversarial Machine Learning in Real-World Computer Vision Systems and Online Challenges (AML-CV), Workshop at CVPR, 2021

 

April

ALFA PhD Candidate, Shashank Srikant's research in collaboration with the MIT-IBM Watson AI Lab is featured in MIT News! Full Article: Toward deep-learning models that can reason about code more like humans

 

March

 

Cognitive Neuroscience Helping AI Understand Computer Programs
Shashank Srikant

MIT Inside Track, Meet the Researcher
MIT Technology Review, EmTech Digital, Cambridge MA

 

 

Generating Adversarial Computer Programs using Optimized Obfuscations
Shashank Srikant

MIT-IBM Watson AI lab, Cambridge MA

 

Enabling Autonomous Cyber Security
Spotlight on Cyber Research: From Lab-to-Market
Una-May O'Reilly
MIT Technology Review in Association with Dartrace
 

CSAIL Alliances Student Spotlight: Shashank Srikant, PhD Candidate, discusses his research in understanding programs and programming. 

 

February 

Artificial Adversarial Intelligence
Una-May O'Reilly
Mongomery Blair High School Computer Science Club, Silver Spring MD, 2021
 

Artificial Adversarial Intelligence
Una-May O'Reilly
Weekeng, IEEE Student Branch, Istanbul Turkey, 2021
 

ALFA welcomes new students to students this semester: Avital Baral, MEng, Helen Landwehr, TPP Lincoln Laboratory Military Fellow, and UROPs: Haimoshri Das, Xinhe Zhou, Ray Dedhia, Evan Rubel. Sasha Shashkov returns as a reserach intern. 

 

January

Getting Adversarial AI to Secure Us From Hackers
Una-May O'Reilly
India Science Festival, 2021
 

We bid farewell to Postdoc Associate, Chathika Gunaratne. Chathika is off on a new adventure and has taken a positon at Oakridge National Lab. 

ALFA says goodbye to PhD Candidate, Bryn Reinstadler. Bryn will be taking up a position at Novartis.

ALFA welcomes new member Andrew Grant, Game & AI Developer.