Minimum Risk Classification
John Tuhao Chen, Lincy Y. Chen, Clement Lee · 2024
This chapter discusses methods of classification that use observable information to predict or estimate an unknown category behind the data. Such classification techniques have various applications, for instance, the diagnosis of a disease, the prediction of up or down trends in the stock market, or the occurrence of a criminal activity in a city, just to name a few. Similar to hypothesis testing problems, for each decision made in classification, there is a chance of correctly classifying the observation into (or not into) a given category, as well as the chance of incorrectly classifying the observation into (or not into) another category.