NEWS
November 2020: Rajitha Hathurusinghe successfully defended his M.Sc. thesis: Building a Personally Identifiable Information Recognizer in a Privacy Preserved Manner using Automated Annotation and Federated Learning.
November 2019: We have received NSERC Engage Grant with Ericsson titled AI Methods for Automated Software Testing
Objectives
The group is performing research in the following areas:
- Merging physical and machine learning models: signal processing, probabilistic machine learning and deep learning models.
- Improving computational efficiency of algorithms: hardware and software accelerators
- Biomedical, tracking, acceleration of software testing and other applications of machine learning and data analysis
Learning in combined phyisical and machine learning models normally relies on computational approximate methods and therefore the term: computational analysis.
History
This research group evolved from Computer Architecture research group (original CARG). Computer architecture research group was formed in 2008 and focused on the following topics: computer architectures, signal processing architectures, hardware and software accelerators, GPU architectures and programming and algorithmic modifications for more efficient implementation in hardware. In recent years our research has gone in several different directions and therefore we decided to change the name of our group to Computational Analysis and Acceleration Research Group (new CARG).
Recent talks and presentations
Rajitha Hathurusinghe, “Building a Personally Identifiable Information Recognizer in a
Privacy Preserved Manner using Automated Annotation and Federated Learning,” University of Ottawa, 2020.
Keynote speech, “Computer Architectures and Algorithms: From Filtering to Deep Learning,” EMCA Developer Conference, Ericsson, Ottawa, May, 2019.