Computer Sciences Dept.

CS 540 - Introduction to Artificial Intelligence

Section 2
Spring 2018

CS 540 Examinations


  • Midterm Examination
    Thursday, March 15, 7:30 p.m. - 9:30 p.m., room 125 Ag Hall
    Closed book. Bring a calculator. No phones allowed. One 8.5 x 11 sheet of paper with notes on both sides allowed.

    Covers topics in first half of the course, including readings, lectures and assignments. That is, covers Search (Chapters 3.1 - 3.6, 4.1, 5.1 - 5.3, 5.5), and some Machine Learning (Chapters 18.1 - 18.4, 18.8.1) The following readings were assigned but will not be covered in the exam: Chapters 1, 2. You are responsible for topics covered in lecture even if there are no readings associated with a topic. Also, you are responsible for topics covered in the readings, except those explicitly excluded, even if they were not covered in class or in the lecture notes. A summary of topics is given HERE.

    • Solutions
    • Average score: 76
    • 90th percentile: 95, 80th percentile: 90, 70th percentile: 85, 60th percentile: 81, 50th percentile: 77, 40th percentile: 74, 30th percentile: 70, 20th percentile: 65
    • Histogram of Scores
    • Approximate correspondence between numeric scores and letter grades:
      	 89-100    A
      	 81-88     AB
      	 72-80     B
      	 65-71     BC
      	 54-64     C
      	 49-53     D
      	  0-48     F

  • Final Examination
    Wednesday, May 9, 12:25 p.m. - 2:25 p.m., room B10 Ingraham Hall
    Closed book. Not cumulative -- covers topics since the midterm exam though will also refer back to basic concepts from the first half of the course. One 8.5 x 11 sheet of paper with notes on both sides allowed. Bring a calculator (not on a phone).

    Emphasizes topics since the Midterm Examination including readings, lectures and assignments. That is, covers readings (Chapters 6.1 - 6.4, 13, 14.1, 14.2, 14.4, 15.1 - 15.3, 18.6.3, 18.6.4, 18.7, 18.9, 18.10, 23.5) and the required papers on deep learning and HMMs. The paper on Bayesian Networks is recommended but not required.

    You should have knowledge sufficient to work through simple examples using the algorithms covered in class including constraint satisfaction problems, neural networks, Perceptrons, backpropagation, deep learning, convolutional neural networks, support vector machines, support vectors, margin, slack variables, kernel trick, basic probability, uncertainty reasoning, full joint probability distribution, marginalization, summing out, conditioning rule, product rule, chain rule, conditionalized version of chain rule, Bayes's rule, conditionalized version of Bayes's rule, independence, conditional independence, Bayesian networks, inference by enumeration, Naive Bayes, Markov model, Hidden Markov model, speech recognition, language model, acoustic model, Adaboost, Viola-Jones face detection algorithm. A summary of topics is given HERE.

    The exam will focus on material since the midterm. Questions may, however, refer back to issues brought up before the midterm, however, so you should refresh your memory about the main ideas and methods from the material associated with the midterm examination. For example, you should be able to relate general search questions to the topics in this part of the course (e.g., what is the search space and what is the search method).

Old Exams

CS 540 | Department of Computer Sciences | University of Wisconsin - Madison