FSS-DSOC Seminar: “Knowledge Discovery and Data Mining in the Social Sciences: Incorporating Machine Learning with Classical Statistical Methods”, by Prof. Xiaoling SHU

Knowledge discovery and data mining (KDD) encompasses a dialectic research process that is deductive and inductive. KDD automatically or semi-automatically considers joint, interactive, and independent predictors to address causal heterogeneity and improve prediction. It plays an important complementary role in revealing valid and significant hidden patterns in data, providing insights into data and theory development, and enriching scientific discovery. The most recent development is to incorporate this new paradigm of predictive modeling with the classical approach of parameter estimation regressions to produce improved models that combine explanation and prediction. I use examples from my research to demonstrate ways to incorporate machine learning with classical statistical approaches to shed light on hidden patterns in data to aid knowledge discovery.

Xiaoling SHU is a Professor of Sociology at the University of California Davis. Her research focuses on the impacts of two of the most profound processes of our times, marketization and globalization, on gender inequalities, subjective sense of well-being, and gender, family, marriage, and sexual behaviors and attitudes. She uses data science models on national and international data to carry out country-specific (China and the United States) and cross-national analyses. Prof. Shu served as the director of East Asian Studies at UC Davis in 2017-22, chair of the Section on Asia and Asian America of the American Sociological Association in 2018-21, president of the International Chinese Sociological Association (formerly NACSA) in 2016-17, and vice-chair and graduate director of Sociology in 2014-17.