From June 27 to 29, the 2026 International Chinese Statistical Association (ICSA) China Conference was held at the Convention Center of the Southern University of Science and Technology (SUSTech). Qikun XUE, president of SUSTech and an Academician of the Chinese Academy of Sciences, Jun ZHAO, Executive Director of ICSA, and more than 800 experts, scholars, and students attended the event.

This conference was jointly organized by the ICSA and the Department of Statistics and Data Science at SUSTech. The theme of the conference was “From Data to Decisions: Statistical Innovations for Complex Challenges,” focusing on the cutting-edge developments, interdisciplinary integration, and innovative applications of statistics and data science in the era of artificial intelligence and big data.

At the opening ceremony, speeches were delivered sequentially by Qikun XUE, Jun ZHAO, Professor George C. Tseng from the University of Pittsburgh, Qiman SHAO, Founding Chair of the Department of Statistics and Data Science at SUSTech, and Xuejun JIANG, Head of the Department of Statistics and Data Science.

In his address, Qikun XUE stated that statistics and data science are key disciplines connecting basic research, technological innovation, and societal applications, playing an increasingly important role in fields such as AI, health, engineering, economics, finance, and social governance. SUSTech places high importance on building foundational and interdisciplinary disciplines and hopes that this conference will further promote international academic exchanges, advance high-level research collaborations, and create a more open academic environment for young scholars and discipline development.

Jun ZHAO, on behalf of the ICSA, congratulated the conference on its opening and thanked the Department of Statistics and Data Science at SUSTech for their organizational efforts. Professor George C. Tseng from the University of Pittsburgh provided a detailed introduction to the structure of the ICSA. The ICSA conferences serve as an important platform connecting Chinese statisticians and data science researchers worldwide, long committed to promoting deep exchanges between statistical theory, methodological innovation, and practical applications.

During the plenary session, Professor Jeremy Taylor from the University of Michigan gave a talk titled “Data Integration: General Concepts and Specific Methods,” introducing how to improve the statistical efficiency of predictive models when individual-level data is limited but external research summary information is available. Professor Annie Qu from UC Santa Barbara presented on the topic of heterogeneity in large-scale, multimodal data with her talk “Representation Retrieval Learning for Heterogeneous Data Integration,” showing how representation learning and machine learning methods can enhance complex data integration and predictive analysis. Professor Niansheng TANG from Yunnan University focused on anomaly detection in incomplete tabular data with his presentation “Variational Masking Generative Model for Anomaly Detection on Incomplete Tabular Data,” introducing a variational masking generative model that learns normal data distribution features and constructs anomaly detection scores to improve model applicability and robustness in scenarios with missing data. Professor Weijie SU from the University of Pennsylvania gave a talk titled “Alignment in Large Language Models: Statistical and Game-Theoretic Perspectives,” analyzing the theoretical foundations and potential limitations of human feedback reinforcement learning from statistical and game-theoretic perspectives, pointing out issues such as preference collapse and difficulty in modeling cyclic preferences in current alignment methods.

The conference provided participants with a multi-level platform for communication, encouraging in-depth dialogue among scholars with diverse research directions, academic backgrounds, and application areas. During the event, attendees showcased research achievements and innovative ideas in statistics and data science, engaged in detailed discussions on the latest trends and developments in the field, and effectively promoted interdisciplinary exchanges and collaboration.