Health Inequality Over The Life Course Between Usa And China: Comparative Evidence From Machine Learning And Conventional Methods
This is an Open Access Thesis.
The United Statas and China are all facing challenges from aging process. Using machine learning methods, conditional Inference Trees and Forests, the thesis estimates the extent to which childhood circumstances contribute to health inequality in old age and evaluates the relative importance of circumstances in the United States and China. The innovative methods draw on a clear-cut algorithm to explain health outcome variability without making strong assumptions on which circumstances play a statistically significant role in shaping health outcomes and how they interact. Also, we will compare the put-of-sample performance between traditional methods and machine learning methods of Inequality of Opportunity analysis. The results support the value of a life course approach in identifying the key determinants of health in old age. Conditional inference tree can be easily mapped the aging population types and thus lay bare the opportunity structure of a given society for a larger audience. On the other hand, conditional inference forests outperform all other methods in terms of out-of-sample performance, delivering the best estimates of the contributions of early life circumstances to health inequality in old age.