Understanding and predicting human behaviour has always been a major challenge due to its inherent complexity and variability. Traditional methods of studying behaviour often occur in controlled laboratory settings that may not accurately reflect real-world scenarios. This lack of real-world context limits the development and application of AI systems, as they struggle to fully comprehend and predict human behaviour. Furthermore, many AI models operate as "black boxes," providing predictions without clear explanations of the underlying decision-making processes, which hinders trust and acceptance.
Behavioural Data Science blends techniques from psychology, economics, sociology, business, computer science, statistics, data-centric engineering, information systems research, and mathematics. This interdisciplinary approach allows us to model, understand, and predict behaviour across three key strands: human behaviour, algorithmic behaviour, and systems behaviour. A significant innovation of our research is the development of human-algorithm systems through the anthropomorphic learning approach. This method integrates behavioural science models with AI algorithms to improve predictions of human, algorithmic, and systems behaviour. It offers more interpretable models, requires smaller training sets, and often outperforms existing deep learning algorithms.