AI and Machine Learning Terminology in Medicine, Psychology, and Social Sciences: Tutorial and Practical...

Area of Science:

• Interdisciplinary research in AI, machine learning, medicine, psychology, and social sciences.

Background:

• Widespread terminological confusion exists regarding AI and machine learning applications.
• Key terms like "prediction" are frequently misused, often applied to studies showing association or retrospective analysis rather than future forecasting.

Purpose of the Study:

• To clarify the correct usage of critical AI and machine learning terminology.
• To provide evidence-based recommendations for consistent terminology to enhance research rigor and public understanding.
• To delineate the hierarchical relationships among AI, machine learning, deep learning, large language models, and generative AI.

Main Methods:

• Review of emerging evidence from systematic reviews on AI and machine learning terminology.
• Analysis of common misuses of terms such as "prediction."
• Explanation of validation, overfitting, and generalization concepts in predictive modeling.

Main Results:

• "Prediction" is often incorrectly used for association or retrospective studies; "prospective prediction" is recommended for future forecasting.
• Essential validation procedures for ensuring model generalizability are discussed.
• Clear distinctions are made between features, independent variables, predictors, risk factors, and causal factors.
• The hierarchical structure of AI and its subfields is clarified.

Conclusions:

• Standardized terminology is crucial for clear communication in AI research across disciplines.
• Accurate use of terms like "prediction" and understanding of validation are vital for advancing AI applications.
• This work aims to foster more rigorous and transparent use of AI in medicine, psychology, and social sciences.