1. Consider the metrics accuracy, precision, and recall.
a. Give one example when accuracy would not be a good performance metric. Give a numerical example.
b. Given one example of a supervised machine learning classification problem when higher precision is desired. Please give a different example than the ones given in class. This need not be a numerical example but must be clearly defined classification problem and dataset.
c. Given one example of a supervised machine learning classification problem when higher recall is desired. Please give a different example than the ones given in class. This need not be a numerical example but must be clearly defined classification problem and dataset.
2. Suppose you are given the same test dataset and two binary classifiers. Give a numerical example such that Classifier 1 has higher accuracy than Classifier 2, but Classifier 2 has both higher precision and higher recall than Classifier 1? Hint: Give a hypothetical 2x2 confusion matrix for each classifier.
