The figure below shows a set of 2-D data samples in two clusters: red for C1 and blue for C2. Could this outcome be produced by performing k-means clustering directly on the 2-D data samples?              

Update the centroids of the two clusters after the above assignment step. m1_new =   ]’,  m2_new = ]’  

Perform cluster assignments in the first iteration, indicating the cluster for each sample in the table below, with 1 for C1 and 2 for C2. x1 x4 x5 x7 x8  

  Problem 5  The following two questions pertain to this problem.

Choose the property below that corresponds to this discriminant function. A. Maximum likelihood classifier B. Minimum error classifier C. Linear classifier  D. All of the above  E. None of the above

Assuming the decision threshold θ = 1.5 for minimum error classification, determine which type of the expected classification error below is larger? A. The error of assigning  

data samples ’ ’ discriminant function g() classification

Problem 7 Answer the following four questions.

Which classifier below resembles a one-layer neural network? A.  Multiclass logistic regression B.  Minimum risk classifier  C.  Minimum error classifier D. None of the above.  

For the same classifier as in the above figure,  if the classification costs are specified as