Richard Harang

TL;DR: In this article, the authors investigate adversarial input sequences for recurrent neural networks processing sequential data, and they show that the classes of algorithms introduced previously to craft adversarial samples misclassified by feed-forward neural networks can be adapted to recurrent Neural Networks.

TL;DR: This work investigates machine learning methods to de-anonymize source code authors of C/C++ using coding style using random forest and abstract syntax tree-based approach, and finds that the code resulting from difficult programming tasks is easier to attribute than easier tasks and skilled programmers are easier to attributes than less skilled programmers.

TL;DR: This paper investigates adversarial input sequences for recurrent neural networks processing sequential data and shows that the classes of algorithms introduced previously to craft adversarial samples misclassified by feed-forward neural networks can be adapted to recurrent Neural networks.

TL;DR: This work states that the introduction of features derived from the Abstract Syntax Tree of source code has recently set new benchmarks in this area, significantly improving over previous work that relied on easily obfuscatable lexical and format features of program source code.

TL;DR: It is shown that programmers who would like to remain anonymous need to take extreme countermeasures to protect their privacy, using both obfuscated binaries, and real-world code found "in the wild" in single-author GitHub repositories and the recently leaked Nulled.IO hacker forum.