A methodological framework for bridging the gap between Artificial Intelligence and Social Sciences Complex cyber-physical-social systems demand rigorous analysis, design, regulation and validation methods that traditional approaches cannot provide. Computational Experiments: A Bridge between Artificial Intelligence and Social Sciences delivers a systematic methodology spanning modeling, simulation, and validation of intelligent systems. Written by leading researchers in complex systems and artificial intelligence, this work provides both theoretical foundations and practical frameworks for studying intricate social and physical systems. The book covers the artificial society modeling framework across four levels: AI agents and prospect theory, learning mechanisms of AI agents, AI society and social networks, and integration with environmental systems. It addresses how computational experiments incorporate generative agents and large language models, and explores policy sandboxes for decision analysis and social system behavior prediction in complex contexts. The book also discusses: Comprehensive coverage of computational experiment methodology including origins, development history, and knowledge frameworks essential for practical applicationsSocial simulation technology foundations providing unique insights into simulating and deducing complex social systems from interdisciplinary research perspectivesDetailed exploration of AI agent architectures incorporating prospect theory, reinforcement learning mechanisms, and multi-agent coordination strategies for system modelingFrameworks for integrating virtual and real-world intelligence to improve predictive capabilities and support decision-making in complex operational environments Designed for professors, researchers, and graduate students in computer science, artificial intelligence, social computing, and systems engineering, this book also serves professionals in policy simulation, strategic planning, and smart system development who require rigorous methods for validating intelligent system behavior.