The CMU Portugal Large Scale Project CAMELOT organized on October 25th its first annual Workshop for an overview and balance of the first half of the project which was launched at the beginning of 2020 and will convene through 2022.
The workshop was held online and counted with representatives from all the institutions involved in the project, including Pedro Bizarro from the project’s leading Company Feedzai, Alcides Fonseca from LASIGE and Faculdade de Ciências da Universidade de Lisboa (FCUL), Bruno Cabral from Universidade de Coimbra (UC), Paolo Romano from Instituto Superior Técnico (IST) and David Garlan from the Computer Science Department at Carnegie Mellon. The Workshop was divided into four (4) sessions, one for each partner Institution starting with Feedzai, who presented work on detecting money laundering and on explaining predictions on time series.
Técnico and CMU followed next with two CMU Portugal Dual Degree Ph.D. students in Software Engineering as speakers. Maria Casimiro started her Ph.D. in 2019 and did an internship at Feedzai during the summer (more about her experience here) and Pedro Mendes who entered the Program in 2021. Both students are being supervised by Paolo Romano (IST) and David Garlan (CMU), researchers of the CAMELOT project, and worked on automatic adaptation and optimization of machine learning systems in the cloud.
In the second session, Alcides Fonseca hosted two presentations from FCUL, one on static verification of Machine Learning pipelines, by MSc student João David, and the second on Genetic Probabilistic Programming by PhD students Guilherme Espada and Paulo Santos, the latter also a PhD Dual Degree student with CMU Portugal.
The final session was chaired by Bruno Cabral, who introduced Qianying Liao, Hugo Matalonga and Francisco Ferreira, from the University of Coimbra, who worked on different approaches for privacy-preserving transfer learning.
The CAMELOT project expects to revolutionize the detection of financial fraud through Machine Learning techniques. The project aims to develop an innovative machine learning platform, which will tackle three key issues that hinder the efficiency and accuracy of modern AI applications such as machine learning models, cloud resources, anonymized data, privacy issues, and integrating information from different, independent, and heterogeneous data platforms.
More about the project here or watch the video below: