Tutorials

The tutorials are in the afternoon of Sunday, May 20th.  The tutorials are free for conference attendees.  

An Introduction to Neural Networks
Sunday, May 20, 2022, 1:30–3:00pm, Grand Salon 1-3

David Bisant, Central Security Service

Neural networks are an important type of machine learning that have generated significant interest due to their success on difficult problems. This tutorial will provide an introduction and a high-level overview of artificial neural networks. It will include the biological basis and the inspiration for these methods, and their colorful history. An explanation of deep learning and deep belief networks, hardware acceleration, and case studies will be used to illustrate how the technology works and the direction in which it is developing.

Large Language Models (LLMs) and Causality Extraction from Text
Sunday, May 20, 2022, 3:30–5:00pm, Grand Salon 1-3

Wlodek Zadrozny, UNC Charlotte

This tutorial delves into how LLMs like BERT, LLAMA, and GPT-3.5/4 can extract causal relationships from diverse text sources. Focusing on business, medical, and newswire domains, participants will learn to identify causes, effects, and actions while navigating challenges of data quality and varying causality definitions. The session bridges practical applications—from building structured representations of technical texts to enabling multimodal question answering—with theoretical foundations, including the mathematics behind model hallucinations and effective prompting strategies. Designed for those with basic machine learning or NLP knowledge, the tutorial provides hands-on experience through example code and resource repositories, offering both practical skills and deeper insights into causal extraction's role in natural language understanding. 

Hands-On Introduction to Quantum Machine Learning
Sunday, May 20, 2022, 1:30-5:00pm, Grand Salon 4

Muhammad Ismail, Tennessee Tech University

Huan-Hsin Tseng, Brookhaven National Laboratory

Samuel Yen-Chi Chen, Wells Fargo

Quantum machine learning (QML) has witnessed great successes in the past couple of years and as such has gained significant attention from academia and industry. However, all the principals of QML are based on quantum information science (QIS). As such, this tutorial aims to present a smooth introduction to QIS and its applications in QML. Further, state-of-the-art QML models will be presented with code examples. This tutorial is meant to encourage academics and students to further explore QML.

Breaking Machine Learning Models with Adversarial Attacks and its Variants
Sunday, May 20, 2022, 1:30-5:00pm, Grand Salon

 Pavan Reddy MS, The George Washington University

Deep Neural Networks are shockingly easy to fool. Small, invisible tweaks to the input can make them misclassify images, misinterpret speech, or even fail in critical applications. This hands-on tutorial dives into black-box and white-box adversarial attacks, revealing how attackers craft imperceptible noise to break ML models. Using Adversarial Lab, an open-source, framework-agnostic tool built for attack experimentation, we will generate and test adversarial examples in real-time, exposing just how fragile machine learning models really are. With live demos, manipulate models yourself, and experience the thrill of outsmarting AI. Think AI is secure? Let’s prove otherwise!