Practice Midterm Exam on Machine Learning and Data Mining | CS 434, Exams of Computer Science

Download Practice Midterm Exam on Machine Learning and Data Mining | CS 434 and more Exams Computer Science in PDF only on Docsity!

1. (32 points) Short questions. The answers should be short (no more than 3 or 4 sentences typically). True/false and multiple choices questions require explainations to your answers. a. (5 points) Define the functional and geometric margins for linear SVM. Explain why is geometric margin a better objective function for learning maximum margin classifier in SVM. J (wtb) 1s the faction wargin . Gemutric margin te fa Shortest Aor bofuoen ths tong pois dad the ceri sibn boundary Function margin cde be axtitrerily sled . Geawtric worn won't change wit Seal tna ty beter | b. (4 points) Explain the difference between lazy learning and eager learning. Given an example of a lazy learning algorithm. Loy lencning - learning stubs blow after 10st exwaple 3 98n, Exoanple ef (ag lenonnd ENN - c. (4 points) (True or False) When learning a LTU with the online perceptron algorithm, no matter what order the training examples are recieved in, the resulting LTU would be the same. False the itu [enmnod B adetetrectrect— epemtl owt on He oderm HE of ‘the. exoanples