From the paper: Predicting Temporal Patterns in Keyword Searches with Recurrent Neural Networks - Phenotyping Human Behaviour from Search Engine Usage
Analysis of psychological normative studies and learning with Deep Learning models.
Constructing manifold-aware adversarial examples using our adaptive neighbourhoods algorithm.
Poster presentation of our work of removing cloud contaminants from ground-based solar imagery.
We demonstrate how, through the development of uniquely-adaptive searchable regions, existing methods can help to further improve the robustness of Deep Learning models, and also make the existing methods applicable to non-image related tasks by providing an upper bound for discovering adversarial examples.
A small presentation given to a select group of students during the summer months. This presentation gives a very brief introduction to Deep Learning and the basic components that make up the feed-forward network.
I present the foundational knowledge for understanding adversarial examples, how we can use the input space to dictate the search space for the existence of these examples, and demonstrate their presence with the use of SAT-solving. This work, as a free and open-source project, provides a framework for ML practitioners to verify their own architectures.
We present novel method to address legal rights for children through a chatbot framework by integrating machine learning, a dialogue graph, and information extraction.