A Psychologically Informed Approach to “Actuarial” Decision Making

Almog Simchon, Michael Gilead*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review


Effective decision making is crucial but often marred by human biases and limitations. Statistical prediction methods have consistently outperformed human judgment, especially in complex and uncertain domains. Recent advancements in machine learning offer further opportunities to improve statistical predictions. While the prospect of human obsolescence arises, we argue that a collaborative approach is still essential. This article reviews recent work emphasizing the integration of human expertise in the development of statistical models that support human judgment. Three key aspects are explored: informed feature extraction, informed priors, and informed data collection. By integrating human expertise, machine learning can produce superior predictive models, allowing for better decision support systems. Collaboration between humans and algorithms remains crucial in leveraging the strengths of both, advancing decision-making capabilities across various domains.

Original languageEnglish
StateAccepted/In press - 2024


FundersFunder number
Ministry of Science and Technology, Israel


    • decision making
    • machine learning
    • natural language processing
    • prediction


    Dive into the research topics of 'A Psychologically Informed Approach to “Actuarial” Decision Making'. Together they form a unique fingerprint.

    Cite this