Use the attached zoo.arff dataset to evaluate the performances of different clustering algorithms. First, observe the dataset. This dataset includes Blank 1 instances (fill in the blank with a number) and represents Blank 2 animal types (fill in the blank with a number). Next, use the EM clusterer and the SimpleKMeans clusterer respectively, and force both clusterers use the same number of clusters as you find from the original dataset. Evaluate and report their performances as follows:

The number of incorrectly clustered instances using EM is Blank 3.

The number of incorrectly clustered instances using SimpleKMeans is Blank 4.