The emerging paradigm of self-structuring AI: theory and applications

Presented by: Professor Damminda Alahakoon and Dr Daswin De Silva Research Centre for Data Analytics and Cognition La Trobe University, Australia. In this seminar, we will present the theory and applications of self-structuring artificial intelligence; an emerging paradigm of artificial intelligence, inspired by human cognition, which aims to address the challenges and complexities of decision-making in non-deterministic data-intensive environments. The theoretical underpinnings of self-structuring AI are composed of learning structures that autonomously evolve, spatially, temporally, laterally and semantically, in alignment with the dynamic, unlabelled, continuously generated nature of data in stochastic environments, process and entities. Incremental learning from evolving data and features, scalable information fusion from high-volume multimodal data, change detection from high-velocity data streams, sequence detection and prediction are some of the new algorithms within this paradigm that we have developed in our research centre. We will also present the application of these algorithms on real-life industry based projects, in smart cities, energy management, digital healthcare, Internet of Things, and social media.