CS 412: An Introduction to Data Warehousing and Data Mining

Course Prefix/Number: CS412

Course Title: Introduction to Data Mining

Graders Needed

Requirement:  Taken CS412 and received A- or higher previously

Job:  Grading Quizzes, Exams, etc. Mostly multi-choice questions.

Course Page:  piazza.com/illinois/spring2019/cs412/home

Instructor: Qi Li

Salary:   $14/hr (less than 6 hr/week)

Contact:  qili5@illinois.edu

If you have friends who may be interested, please let them know as well!

--------------------------------------

Graders found. Thank you very much!

Dear students,

The score of project and review are available on Compass2g. Also, we send the summary of the reviews to you for your reference.

Thanks,

TAs.

Please check the score on Compass2g. The answer is under Resources.

Here are some common mistakes:

1b: missing reasoning / wrong reasoning: -3

1c: DAG is not an ensemble methods -4; (boosting or bagging are correct.) 

1d: Accuracy is wrong for imbalanced data. -4

3 (c) Wrong / missing merging procedure / only code without any explanation -5; two clusters are left as final result -2 (K=2 is the condition of 3(a) not for the whole question.)

-TA

Dear Students,

Thank you for taking this course and your hardworking this semester. I hope you all learned something useful from the classes.

Here are some final announcements.

1. I will take off the course material (homework and exam) from Piazza and inactivate the course this Sunday. After inactivation, you will not be able to post on the website. If you have questions after that, please email me. 

2. For students who are assigned with report reviews, the deadline is tomorrow (12/20/2018) noon on Compass2g. Please submit it on time. 

3. The TA team is done with the final exam grading and you should see it on compass2g now. The mean is 68.96 and the median is 72, with std=17.91, and the grade distribution is as follows

The final exam will not be handed back and you will not be able to review it. But you can still request a regrade if you believe your score is wrong. However, to avoid request just for luck, we use the following policy:

-10 points if your score does not change after regrading. 

That is, if your score is 76, and you request a regrading. We will regrade your entire paper. If there is no mistake, your final score will be 76-10=66. If there is a mistake, then your score will be updated to the new score without the -10 (note: the new score can be higher or lower). 

If you want to request for regrading, please do so by this Friday. 

The rough rubric is as follows.

For questions worth 1 or 2 points, no partial credits. 

For the question on measures, each slot is 1 point

For the question on pattern constraint, each constraint worth 3 points. Partial credit is given if your answer is incomplete.

For the 4 big calculation questions (decision tree, naive Bayesian, k-median, and agnes), you will get most of the points if your procedure is correct and you use the right formulas. The final answer is 1 point.

For the bonus questions, partial credit is given. 2(b) might be the most challenging one, and 5 students got it right. 

4. The TA team is still grading HW5 and project.

5. The guideline for letter grades in @273 may be changed minorly (2-ish points) since we added a bonus assignment. The major change is the score for A+. Since A+ is a grade of honor, it will be rewarded to the top 10% in the class. The 3-credit hour and 4-credit hour will have a different cut-off to ensure fairness.

Hope you all have a nice winter break! Happy Holiday!

Best,

Qi

Please check the following google doc to see the coverage of the exam.

https://docs.google.com/document/d/1-ubMImBCpPU50tkyJ48xdudp5b5XoJz--lEwpajbTuQ/edit?usp=sharing

The teaching staff has posted a new homework resource.

Title: hw5-answer.pdf
http://www.piazza.com/class_profile/get_resource/jkiigzqdvsmay/jpn3pj2gi6c1ka

You can view it on the course page: https://piazza.com/illinois/fall2018/cs412/resources

#pin

Since many students requested the acceptance of late submission of HW5, but many students also ask for the release of solution, we will adopt the following late policy:

10% off for one day (Dec. 12th, 11:59 PM),
20% off for one and half days (Dec. 13th, NOON, 11:59 AM),

The solution will be released on Thursday early afternoon on Piazza. Any submission after that will not be graded.

#pin

Dear Students,

We decide to extend the deadline for the report reviews. The new deadline is Dec 20, noon. Please note that this is a hard deadline, as we also have a hard deadline to upload your letter grades. If you do not submit your reviews in time, you will affect the grades for 2 groups, and we won't have enough time to fix it. 

Qi