Machine Learning

10-601, Fall 2011

Carnegie Mellon University

Tom Mitchell, Aarti Singh


Homework 1: pdf ,   Code and Data,    Tex Source,    Solutions.

Homework 1 Corrections and Clarifications:

  1. While this correction pertains to an already completed homework, it is important to note there was an error in question 2.4. - the alternative splitting method. The question stated that if you minimized the equation for O(a,t) with respect to a and t, you would find the optimal split for the misclassification rate criteria. However, this function was missing something important. The terms summing the number of samples misclassified above and below the split point should have been normalized. Specifically, the term summing the number of samples misclassified above the split should have been divided by the total number of samples above the split and the term summing the number of samples misclassified below the split should have been divided by the total number of samples below the split.

Homework 2: pdf ,    Data,    Tex Source,    Solutions.

Homework 2 Corrections and Clarifications:

  1. The original homework assignment stated there was a third optional question. This was incorrect. There are only two required (and no optional) questions.
  2. When using the MAP estimate for question 2.5, note that hallucinating each word appearing Beta times in the training set corresponds to having a Dirichlet prior with all parameters equal to (Beta + 1). Ensure that you are not adding (Beta - 1) to your word counts, since that may result in negative probabilities. Also, we are hallucinating each word appearing Beta times in the entirety of the training set, not Beta times per document.

Homework 3: pdf ,    Tex Source,    Solutions.

Homework 3 Corrections and Clarifications:

  1. Question 3.1(a) The joint distribution on the left of the equation in 3.1(a) should be P(A,B,C,D,E,F,G) and not P(A,B,C,D,E,F).
  2. Question 2(d): Please replace the current hint by this hint: "Your expression can involve integrals, and if it does you need not solve them. However, there is a much simpler answer that does not involve integrals so look for that!"

Homework 4: pdf ,    Tex Source,    Solutions.

 

Homework 5: pdf ,    Tex Source,    Solutions.

Homework 5 Corrections and Clarifications:

  1. Question 3.1 This question incorrectly sometimes indexed the vector x by t. However, these vectors should have been indexed by i. The homework file has been updated

Homework 6: pdf ,    Tex Source,    Solutions.

 

 

Format- Submitted homeworks may be either typed or handwritten. However, for ease of grading, please submit answers to the individual questions that make up each homework assignment on separate pieces of paper. When turning in code, please both print and attach a copy of your code to your homework and submit your code through the course blackboard website.
Note - We might reuse problem set questions from previous years, covered by papers and webpages, we expect the students not to copy, refer to, or look at the solutions in preparing their answers. Since this is a graduate class, we expect students to want to learn and not google for answers.

Collaboration policy - Homeworks must be done individually, except where otherwise noted in the assignments. 'Individually' means each student must hand in their own answers, and each student must write their own code in the programming part of the assignment. It is acceptable, however, for students to collaborate in figuring out answers and helping each other solve the problems. We will be assuming that, as participants in a graduate course, you will be taking the responsibility to make sure you personally understand the solution to any work arising from such a collaboration. Students who collaborate in this way are expected to list the name of those they collaborate with in their homework submissions.

Late homework policy -
  • Late homeworks will be penalized according to the following policy:
    • Homework is worth full credit at the beginning of class on the due date.
    • It is worth half credit for the next 48 hours.
    • It is worth zero credit after that.
  • Turn in hardcopies of all late homework assignments to Sharon Cavlovich. Put down the date and time of submission on the HW sheets when submitting your assignments to Sharon. If she is not available, please slide your HW under her door.
  • Don't forget that your lowest homework score will be dropped from your final grade.