One common problem in implementing statistical algorithms is being able to sample from an arbitrary distribution: suppose you have N objects, each with a different probability of being chosen - how do you sample these objects? We apply the solution to sampling from a weighted dataset to get unbiased estimates of the original dataset.

Learn:

• What a weighted survey data set means.
• Analyze data from the National Survey of Family Growth.
• Geometrically understand how to sample from discrete distributions.