Question 1:

Consider a binary classifier that classifies a share into Buy or No-Buy in a trading algorithm. The classifier was experimented on training data composed of 1000 manually labelled shares. 500 shares are labelled as Buy and 500 shares are labelled as No-Buy. This constitutes the ground truth data. The classifier provided the following output for the 500 shares manually labelled as Buy:

400 shares classified as Buy and 100 shares classified as No-Buy. The classifier provided the following output for the 500 shares manually labelled as No-Buy:

490 shares classified as No-Buy and 10 shares classified as Buy.

1. Calculate the Accuracy, Precision, Recall and F-measure.