Decentralized AI and learning from private data

Research in an academic environment is a chance to solve real-life problems and imagine the technology of the future. My current research is on methods and algorithms for decentralized artificial intelligence, machine learning and data mining. Computer-generated data is now everywhere, from business to medical and scientific data visualization, and despite major advances in machine learning and AI, many problems still limit how effectively it can be used.

My main research goal is to create tools and algorithms for efficient machine learning and AI with private data. Data regulated under GDPR and HIPAA is increasingly complex to analyse, yet understanding it is essential, both to serve people better and to use that knowledge for decision-making. While preserving privacy, I develop methods for knowledge discovery, machine learning and AI applications that decision-makers can rely on.

During my Ph.D. I applied knowledge discovery, big data, machine learning and multiple criteria decision analysis to problems such as road safety, and proposed methods to evaluate association rules from the KDD process that have been used in road safety, text mining and social networks. I also bring substantial industry experience in data mining and applied machine learning.

In the coming years I will explore machine learning and federated learning further, together with researchers in computer science and in fields such as healthcare and business. This agenda needs technical innovation along three dimensions: models, algorithms and experimental evaluation.