Human-AI interaction in predictive systems: Enhancing AI predictive systems with expert knowledge and human-centered explanations

For the effective adoption of predictive AI systems in industry, it is crucial to optimize the human-AI interaction. An improved interaction with advanced predictive AI systems can enable human experts to enhance the accuracy and efficiency of their decisions and thus bring additional value to their company.
The primary objective of this project is to investigate how various 
interaction techniques, (such as explanations generated by LLM’s) influence the adoption of these AI models and impact the accuracy and efficiency of human experts working with them.
The secondary objective of this project is to explore how we can harness human expert knowledge to improve prediction models.
To achieve these goals, we will develop a user-friendly model agnostic platform that enables different interaction methods and facilitates the study of how users interact with these various methods and different predictive AI models.