A micro-service reference test application for model extraction, cloud management, energy efficiency, power prediction, single- and multi-tier auto-scaling

General-purpose library for extracting interpretable models from Multi-Agent Reinforcement Learning systems

Analytic tableau based minimal model generator, model checker and theorem prover for first-order logic with modal extensions

MEME: Generating RNN Model Explanations via Model Extraction

Simple machine learning in Python/Tensorflow with model saving

For our AAAI23 paper "DisGUIDE: Disagreement-Guided Data-Free Model Extraction" (Oral Presentation) by Jonathan Rosenthal, Eric Enouen, Hung Viet Pham, and Lin Tan.

Ecore metamodel reverse engineering: Automatically extract EMF metamodels from Java code.

📄 [Talk] OFFZONE 2022 / ODS Data Halloween 2022: Black-box attacks on ML models + with use of open-source tools

CME: Concept-based Model Extraction

Marich is a model-agnostic extraction algorithm. It uses a public data to query a private model, aggregates the predicted labels, and construct a distributionall equivalent/max-information leaking extracted model.

The Labelled Transition Systems Extractor tool project

Graphical User Interface to debug ROS systems

Code for VidModEx: Interpretable and Efficient Black Box Model Extraction for High-Dimensional Spaces

Serverless Application Extraction System (SAES)

🚀 A FastAPI-powered project that uses a trained ML model to classify incoming emails as spam 🛑 or ham 📧 in real-time. Supports CSV batch processing 📂, multi-query input 🔄, and detailed logging with usage stats 📊.

Extension for the LTS Extractor platform, which is used for enabling the communication with the methods for generating model analysis, and to ease the management of log files throughout the platform.

Collection of the TeX files and figures used to create my UofA CS master's thesis