Statistical Data Analytics for Decision-Making

The problems included in this section illustrate a close connection of management engineering applications with some methods of advanced statistical data analysis. Five problems are included: (i) principal component decomposition of the population demographic dataset followed by regression with principal components to identify the demographic variables that most contribute to the hospital operation margin, (ii) cluster analysis to identify the distinct groups of ZIP codes (clusters) in terms of patient contribution margin, (iii) forecasting of the patient census time series using recursive linear digital filtering technique, (iv) Bayesian inference for interpreting results of single and multiple infection diagnostic tests, and (v) the expected number of patients discharged from emergency department using the concept of conditional probability of patient discharge from ED.