QBUS3820 Lecture Notes - Lecture 3: Model Selection, Test Statistic, Mahalanobis Distance

• School: University of Sydney
• Department: Business Analytics
• Course: QBUS3820
• Professor: Peter

Document Summary

Discipline of business analytics, the university of sydney business school. Lecture 3: linear regression and k-nearest neighbours: statistical properties of ols, the gaussian mlr model, k-nearest neighbours, comparison with linear regression. In classical statistics, the population parameter is xed and the data is a random sample from the population. We estimate by applying an estimator b to data (in our case the ols estimator). As the sample on which estimator b is computed is random, the value of the estimator is a random variable, and the distribution of this random variable is called the sampling distribution. We can study the uncertainty of an estimate by understanding the sampling distribution of the estimator. Suppose that we draw a large number of di erent datasets d(s) (s = 1, . On each of these datasets, we apply the estimator b and obtain a set of estimates {b (d(s))}s thought of as the distribution of these estimates as we let s . s=1.