HiPerCog group's core funding comes from the Academy of Finland, as a part of the University of Helsinki profiling area 'Mind & Matter - Foundations of Information, Intelligence and Consciousness'.
This profiling area considers how interacting with technology affects humans, and how to develop human-friendly AI technologies, by understanding the subjective, objective, and experiential nature of human intelligence, consciousness, and their neural underpinnings.
HiPerCog's Mind & Matter subproject, with funding for one post-doc, one PhD student, and lab development, reflects this with a focus on human learning using computational cognitive neuroscience methods.
The project GUESSED (Grappling with Uncertainty in Environments Signaling Spurious Experiential Decisions) employs one post-doc and one PhD student. GUESSED is funded by Nordforsk, with partners from Norway and Sweden, including Arctic University of Tromsø, and University of Umeå. The project is led by Center of Avalanche Research and Education.
In GUESSED we study decision making in a highly-demanding natural environment, avalanche terrain. We work with world-leading experts to elicit their expertise and create ML models to help non-experts structure their decision making process as an expert would.
HiPerCog studies electroencephalographic (EEG) neural correlates of cognitive attention deficits in adults diagnosed with attention-deficit/hyperactivity disorder (ADHD); and aims to translate the findings to evidence-based neurofeedback (NFB) treatment.
It is important to establish falsifiable hypotheses of the neurocognitive mechanisms specific and unique to ADHD, which can help us identify biomarkers to supplement current subjective diagnostic practice. Knowledge of neurocognitive mechanisms of ADHD can also help develop alternatives to pharmacological treatment, such as NFB.
The AI Personality and Cognition (AiPerCog): AI learning from Humans in Games is funded by the Research Council of Finland (#355200). AiPerCog aims to find behavioural markers of psychological states, such as negative or positive reinforcement from play, based on computer gaming behaviour. To do this we will: (1) data-mine very large datasets of game replays to identify human patterns of play, (2) train AI agent models to replicate these patterns, & (3) build classifiers to link psychological profiles to play patterns. We apply this methodology to game-play records of popular games to understand how different types of reinforcement from game situations drive human player behaviour.