There will be seven homework assignments, each due at least one week after it is handed out. Homeworks will explore applications of statistical machine learning using Python, supported by some mathematical derivations and analysis. Homeworks determine 60% of your overall grade. We will drop your lowest homework score, and average the other 6 scores equally, to determine an overall homework score.

Python

In CS275P this year, we will use Python 3 for the programming portions of the homework assignments. We highly recommend installing the Anaconda Python distribution, which includes the full SciPy stack as well as a number of other useful tools. We recommend using either the Spyder development environment or Jupyter notebooks for your course assignments. Both are included with Anaconda and can be launched from the Anaconda Navigator application. If you are on a computer without Anaconda installed, you could also use a Google Colab notebook, or create a Jupyter notebook on the ICS JupyterLab server (VPN required).

Spyder is designed specifically for scientific computing with Python and natively supports plotting with the MatPlotLib library. It is an intuitive option if you are used to IDEs like MATLAB, RStudio, or Eclipse.

Jupyter notebooks provide a convenient way to integrate all required parts of your homework solutions. A single notebook can include text, mathematical equations in LaTeX, figures, Python code, and the output of executing that code. Notebooks can be exported or printed as PDF files (check that all content is visible!) for submission to gradescope. However, it may take a bit of extra work to understand how to setup and use Jupyter notebooks. To export a PDF file for online submission, we recommend that you:

From the Jupyter menus, choose "File > Print Preview". This should produce a new webpage where all of your answers are visible.
Once you've checked your answers, from your browser menu (we use Chrome) choose "File > Print" (or a similar command) and then print to a pdf.
Verify that your printed pdf is multiple pages, not one giant page.

This approach is easier to setup, and tends to work better, than using internal Jupyter commands/menus to export directly to pdf.

Collaboration and Academic Honesty Policy

Students are encouraged to discuss and work on homework problems in groups. However, each student must write up their solutions independently, and do any required programming independently. You may not directly copy solutions or code from other students, or from materials distributed in previous versions of this or other courses. You may not make your solutions available to other students: files in your home directory may not be world-readable, and you may not post your solutions to public websites or repositories.

Academic dishonesty is unacceptable and has serious consequences at the University of California, Irvine. It is each student's responsibility to know and follow these guidelines, as well as the UCI ICS School Policy on Academic Honesty.

Formatting Homework Solutions

Your answers to all questions, including all Python source code, must be included in a single pdf document. Number your answers by question and part using the same order as in the handout. Result plots must be submitted as figures in the same pdf as your other answers, not as separate files.

If you'd like, you may integrate your Python code with the other parts of your answers. It is also fine to include all Python code at the end of the pdf, after your other answers.

Be sure to show your work! You may lose points if you only report the final numeric answer with no indication of how you computed it. Also even if your final answer is incorrect, you will receive more partial credit if your derivations and reasoning are clear.

The graders will look at your code to assess correctness and give partial credit, but they will not execute it. Even if a question involves programming, we should be able to find your answers to all parts in the solution pdf, not solely in the source code. If we ask you to report a number that your code computes, it should go in the solution pdf. Code comments are encouraged to improve readability and show your work, but should not be used to report the result of executing that code.

We recommend that you use the LaTeX to typeset your solutions, but if you prefer, you may use another program that generates easy-to-read pdf files. To learn more about LaTeX:

Wikibooks LaTeX Manual, especially the sections on mathematics, lists, importing graphics, and floats and figures.
LaTeX tutorial videos prepared by Joe DiNoto.
MacTeX and MiKTeX make it easy to install integrated LaTeX environments on your Mac or Windows computer.
Overleaf provides an online, graphical LaTeX development environment.

Any English text in your answers should be typed when possible, not handwritten. We allow an exception for mathematical derivations, which if you prefer may be scanned and included in your single solution pdf. However it is your responsibility to make sure that any scanned math is legible, and graders have the discretion to deduct points for illegibility. Remember to show your work clearly, and make it easy for the graders to check your work!

Submitting Homework Solutions

To submit your homework for grading you must upload to gradescope, not Canvas. Gradescope requires your solutions to be contained in a single pdf, so you must export or print to pdf from whatever program(s) you use to prepare your answers. Note that many pdf viewers include an option to concatenate multiple pdf documents into a single, longer document.

Late Submission Policy

Homework assignments are due by 11:59pm on the posted due date. Your answers may be submitted up to 7 days late (by 11:59pm one week after the deadline); after this point, solutions will be distributed and handins will no longer be accepted. We do not deduct points for submissions that are (less than 7 days) late. But to encourage you to keep up with the material presented in lecture, for each homework that is submitted on time, you will receive 1 bonus point (out of a maximum of 100) on your final project.

Exceptions to this policy are only given in very unusual circumstances, such as documented family or medical emergencies, and any extensions must be requested in advance by e-mail to the instructor.

Homework Grading

Homework grades will be distributed electronically via gradescope, with a rubric indicating why points were deducted. If you would like to dispute the points you received for some assignment, do so by using the regrade request function on gradescope. We will only allow regrade requests for one week after the date that grades are returned.

Resolving regrade requests may require discussion among the course staff, so please be patient. Certainly do tell us about cases where a correct answer was missed or given only partial credit. If your answer is partly correct and you received partial credit, we are unlikely to change the amount of partial credit.

We will only consider regrade requests submitted via gradescope, and the course staff will not discuss homework grading during office hours. If you have significant concerns about your overall course standing, please arrange an individual meeting with the instructor.