Course Details

CS 6220: Data Mining Techniques

Special thanks to Professor Virgil Pavlu to use the materials from his previous classes. This course has been developed to incorprate those materials, along with the latest advancements in data sicence and machine learning for data mining.

Schedule: Wed 6:00pm - 9:20pm

Location: Zoom + SH 325

Dates: Sep 4, 2024 - Dec 14, 2024

Instructor: Jin Yu | jin1.yu@northeastern.edu | Office Hours: Zoom Meetings (see Canvas Course Syllabus for schedule)

Piazza: Questions and lecture material are handled via Piazza | Sign up https://piazza.com/northeastern/fall2024/cs6220yu

Canvas: Course schedule and assignments are available via Canvas | Log in at https://northeastern.instructure.com/courses/196663

Teams: Office hours, Friday 11:30-12:30, are held virtually via Zoom Meetings | Please join Professor Jin Yu's Office Hours (Zoom Meeting)

Recommended Textbooks:

  • Data Mining: The Textbook. Charu C. Aggarwal, Springer, 2015.
  • Mining of Massive Dataset, 3rd Edition. Jure Leskovec, Anand Rajaraman, Jeff Ullman, Cambridge University Press, 2020.

Prerequisites:

  • Linear algebra tutorial: http://www.stanford.edu/class/cs229/section/cs229-linalg.pdf
  • Probability tutorial: http://www.stanford.edu/class/cs229/section/cs229-prob.pdf
  • Programming: We will mainly use Python, you should ideally have completed DS 5010 or equivalent.

Academic Integrity:

Be familiar with the university’s academic integrity policy on cheating and plagiarism.

Overview

This course delves into the core algorithms and inference techniques fundamental to data mining. Students will explore various methods for data analysis, including classification, clustering, association analysis, and anomaly detection. Through practical examples and hands-on projects, the course aims to equip students with the skills to apply data mining algorithms effectively to large data sets and derive insights for decision-making. Emphasis is placed on understanding the algorithms' mechanisms, their applications, and their impact on real-world problems.

Course Progression Requirement

Students are required to get a B or above in the placement courses in order to progress into the core courses in the degree program. Students that do not achieve a B or better in the placement courses will be required to retake the courses.

Topics

  • AI fundamentals
  • Data mining basics
  • MapReduce and Spark
  • Senstive hashing
  • Clustering(KMeans, DBScan, etc.)
  • Dimensionality reduction(PCA, tSNE, etc.)
  • Feature analysis and selection(SHAP)
  • Classification methods(Neural networks, decision trees, etc.)
  • Variational methods
  • Sampling methods
  • Hidden Markov models
  • Data fusion
  • Graphical models
  • Causal inference

General Policies

Name and pronouns

Please let me know if you use a different name or pronouns from what appears the class roster. You may use a chosen name on Piazza and when submitting assignments and exams, but please be consistent and inform the instructors. The Northeastern LGBTQA Center can provide resources for changing your name and gender marker in the Northeastern system.

Please be kind and respectful to your fellow students regardless of identity or background. Students are expected to respect and use other students’ chosen names and pronouns. All students are expected to respect Northeastern’s commitment to diversity and inclusion.

Mental and physical health

Please reach out to me as early as possible if you have difficulty keeping up with class material or completing assignments for personal reasons. I am able to provide more accommodations and options for you earlier in the semester than later in the semester when deadlines are looming. The We Care program at Northeastern University is another resource available to you in times of stress.

Academic integrity

All students are expected to abide by the university’s academic integrity policy. Plagiarized work will not receive points in this course and may be reported. Authorized use of outside resources (including but not limited to third-party code) must be cited.

Title IX

Northeastern University strictly prohibits discrimination or harassment on the basis of race, color, religion, religious creed, genetic information, sex, gender identity, sexual orientation, age, national origin, ancestry, veteran, or disability status. Please review Northeastern’s Title IX policy, which protects individuals from sex or gender-based discrimination, including discrimination based on gender-identity. Faculty members are required to report all allegations of sex/gender-based discrimination to the Title IX coordinator.

Remote Instruction

IMPORTANT: I might be teaching remotely when it snows. Please see Canvas for Zoom links, and check Piazza for updates.

Students may participate remotely via online Zoom meetings. All course content can be accessed and completed remotely. However, some content and assignments require synchronous attendance (i.e., during the regularly scheduled class time in the Boston time zone), such as quizzes and project presentations. Students are still responsible for making sure they satisfy any college requirements for in-person enrollment.

The instructor may teach some class sessions fully remotely if the need arises.

Please do not come to class in-person if you are experiencing symptoms of COVID-19 or other flu-like illness.

Technology

Piazza

Course administration, including all questions, course materials, and course announcements will be handled via Piazza.

Please do not email instructors or TAs directly – use Piazza for your questions and queries instead. This allows us to track all course-related correspondence in a single location.

General questions that may be useful to other students should be posted publicly to the whole class. If your question is specific to you, or includes a partial solution, then post it privately to instructors only.

Please see this Stackoverflow guide for how to ask a good question.

Canvas

Assignments, quizzes, and grading will be administered via Canvas.

All assignments and quizzes will be posted on Canvas, and must be submitted on Canvas by the posted due date. Please do not email completed assignments or quizzes to instructors or TAs, or post them on Piazza.

Zoom

Classes will be broadcast synchronously via Zoom. Students can use Zoom to attend class virtually when in-person attendance is not possible.

Microsoft Teams

Virtual office hours will be held via Zoom Meetings. During scheduled office hours or by appointment, instructors and TAs will be available for live chat or video call on Microsoft Teams. You will be automatically added to a team for the course.

The instructor will hold office hours in the “General” channel, and TAs will hold office hours in the “TA Office” channel. The schedule for office hours can be found on Piazza Resources under the “Staff” tab.

Assignments

Homework

Homework assignments will be assigned every 2 weeks and must be completed individually. Each homework is due online via Canvas on the date scheduled on Canvas. (Please refer to the deadline on the actual assignment rather than the semester schedule, which is tentative and may not be updated if the assignment changes.)

Some aspects of the homework may be discussed with each other, but they should be completed individually, and your submitted work should be your own. Sharing of worked solutions will not be tolerated and will be considered cheating. Plagiarized solutions will receive a zero. Solutions with a very high degree of similarity with another past or current student’s will be considered plagiarism, and will be treated accordingly.

Labs

This lab involves using Jupyter Notebooks to implement and explore various data mining algorithms effectively. All labs must be completed individually and each lab is due online via Canvas on the dates scheduled on Canvas.

Exams

There will be midterm exam and final exam. Both exams will be completed online via Canvas in the classroom on the dates scheduled on Canvas, and will replace one class meeting.

Late work and grading

Late assignments will not typically be accepted. No-penalty extensions may be given on a case-by-case basis if requested at least 48 hours in advance of the due date with a reasonable justification.

Petitions for re-grades must be made in writing via Piazza private message to the instructor no later than 1 week after receiving the original grade. The petition must clearly explain why a re-grading is justified and why your answer(s) should be considered correct. The new grade may be lower than the original grade.

Before petitioning the instructor for a re-grade, students should first contact the grader to make sure they understand why they lost points.

Grade Scale

Students are required to get a B or above in the placement courses in order to progress into the core courses in the degree program. Students that do not achieve a B or better in the placement courses will be required to retake the courses.

The grade in this class is distributed as follows:

  • Homework: 30%
  • Labs: 40%
  • Exams: 30%
  • Extra Credit: up to 2%

Final grades will follow the following scale:

  • A : <= 100.0
  • A-: < 94.0
  • B+: < 90.0
  • B : < 87.0
  • B-: < 84.0
  • C+: < 80.0
  • C : < 77.0
  • C-: < 74.0

These scales are subject to change at the discretion of the instructor.