Exam Summary

Overall the grades were quite good!

• 33-35: 35, 35, 35, 34, 34, 34, 34, 33.5, 33.5, 33, 33, 33, 33 (14 scores)

• 30-32: 32, 32, 31.5, 31.5, 31.5, 31, 31, 31, 30.5 (9 scores)

• 26-29: 29, 28.5, 28, 28, 27, 26.5, 26, 26, (8 scores)

• (0-25): 24, 20.5, 20 (3 scores)

Remarks

• The key distinction between supervised and unsupervised learning is whether one has labeled datasets. Many people got perfect scores on Problem 1, but some reversed “supervised” and “unsupervised”

• Imputation is a technique for handling missing data within a dataset. It involves not discarding data (the opposite), and in general, we don’t want to discard data whenever possible.

• Some people got confused about the definition and/or mathematical notation associated with the perceptron.