COS 302 - Mathematics for Numerical Computing and Machine Learning

Fall 2020

Course home Outline and Lectures Assignments


Exams

Final Exam - Sun, Dec 13, 1:30 PM EST

The final exam will be short answer / multiple choice, and will be taken online in Gradescope. It will be designed to be doable in 60 minutes, but you will have a block of 180 contiguous minutes to complete it. The time officially scheduled by the registrar is 1:30-4:30 PM on Sunday, December 13, 2020. But the registrar has given course instructors the authority to "provide exam time adjustments for students who have time zone complications". The way this will be accommodated in COS 302 is that you may complete the exam at any point within the 24-hour block from Sunday 1:30 PM EST through Monday 1:30 PM EST. There is no need to notify the course staff if you are taking the exam any time within that block.

Topics will include multivariable differentiation, probability, and optimization. Material in your textbook that was not mentioned in lecture or precept will not be on the exam. In addition, some of the details presented in lectures but not in the textbook will not be on the exam - this specifically includes probability distributions including Poisson, Gamma, and Beta, inequalities such as Markov's and Chebyshev's, and the concepts of entropy and mutual information. Material from assignments and precepts is fair game, though.

You are responsible for topics including:

The focus will be on concepts, not on hand-calculation - you should expect a style similar to the midterm exam. You will be expected to be able to read Python code, but will not be asked to write Python. (Though it's possible that you may be asked to explain a technique by writing a short segment of pseudocode.)

During the exam you MUST NOT refer to the textbook, course materials, notes, or any information on the Internet. The only exception is that you MAY prepare a one-page, one-side, handwritten set of notes, and you should have blank scratch paper available. You will be required to type out and sign an honor code pledge that you did not use any unauthorized sources of information during the exam.


Final Exam Practice Questions


Final Exam Q&A

There will be a Q&A on Monday, Nov 23 during the regular class time. Please come prepared with questions, either about the practice questions above, or related to any topic we've covered so far.



Midterm Exam - Wed, Oct 7

The midterm will be short answer / multiple choice, and will be taken online in Gradescope. It will be designed to be doable in 45 minutes, but you will have a block of 90 contiguous minutes to complete it, at any point within the 24-hour block from Wednesday noon EDT through Thursday noon EDT.

Topics will include all of the linear algebra concepts we've covered so far, up to and including the lectures and precepts of the week of Sep 28. Material in your textbook that was not mentioned in lecture or precept will not be on the exam. Material from assignments and precepts is fair game, though. Topics include:

The focus will be on concepts, not on hand-calculation. You will be expected to be able to read Python code, but will not be asked to write Python. (Though it's possible that you may be asked to explain a technique by writing a short segment of pseudocode.)

During the exam you MUST NOT refer to the textbook, course materials, notes, or any information on the Internet. The only exception is that you MAY prepare a one-page, one-side, handwritten set of notes, and you should have blank scratch paper available. You will be required to type out and sign an honor code pledge that you did not use any unauthorized sources of information during the exam.


Midterm Exam Practice Questions

Expect less of a focus on hand-calculation than in those questions. Because of the online format this year, expect more short-answer and multiple-choice questions.


Midterm Exam Q&A

There will be a Q&A on Monday, Oct 5 during the regular class time. Please come prepared with questions, either about the practice questions above, or related to any topic we've covered so far.



Last update 18-Nov-2020 12:09:55
smr at princeton edu