Student performance for the course will be evaluated via tests (done individually), assignments (done in pairs or individually), peer grading, and Kaggle competitions.

Due to the huge number of students and the subjective nature of many of the results, this term I’ve decided to try turning evaluation of assignments into a learning experience for students. We will be using Kritik, a peer-to-peer learning and evaluation platform, read here for more about how Kritik will be used for Peer Grading in this course.

Assignments Logistics

  • Collaboration: assignments can be done in pairs or alone
  • Topics: Assignments will arise from the major component topics of the course, some will buid on previous assignment outcomes.
  • Possibility to have later assignments as Kaggle-style competitive submissions (note: vast majority of grade will be based on performance and correctness rather than based on competitive performance)
  • Completing the assignments will require multiple skills:
    • mathematical analysis of data and results
    • logical design and clear description of a expeirmental methodology
    • programming various algorithms for processing, training and analysis of data to achieve given tasks (programming will be in Python using libraries such as sci-kit learn and tensor flow)
  • Note that the assignments will be shorter, less questions, and more frequent than in previous iterations of the course.
  • Grading will be done by Peer Review via Kritik

Information and Rules for all assignments:

  • Assignments can be done alone or by a pair of students.
  • Assignment questions and approaches can be discussed online, people can help others find the right approach and give hints. But no code will be passed between groups, and people should refrain from spelling out exactly what to do at the lowest level of detail.
  • You can read online resources about the question to help answer it if you do not understand enough to do so, or if the slides and lectures have not succeeded in helping you understand how to answer it.
  • Late Assignments:
    • Assignment due dates will be at midnight on the designated date.
    • Late assignments will have the following penalties from the assigned grade:
      • 6 hours (0%)
      • 6-24 hours (5%)
      • 24-48 hours (10%)
      • >48 hours (100%)
    • If you know ahead of time that you will not be able to make the deadline do to serious health or personal issues contact the professor to ask for an exception.

Assignment Weighting and Topics (New, as of Feb 21, 2022)

  • Asg 1 - Weight: 10% - Topics: Data Preprocessing, Basic Classification with KNN
  • Asg 2 - Weight: 15% + 5%=20% - Topics: Representation Learning, Parameter Estimation, Naive Bayes, Classification and Regression with Decision Trees and Ensemble Methods
  • Asg 3 - Weight: 20% + 10%=30% - Topics: Deep Learning for Classification and Regression on Numerical and Image Data
  • Midterm Test - no change 10%
  • Final Exams - Weight 25% + 5% = 30%

(old) Assignment Weighting (Original Description)

  • Asg 1 - Weight: 10% - Topics: Part I - Data Preprocessing, Basic Classification with KNN
  • Asg 2 - Weight: 15% - Topics: Part I and II - Representation Learning, Parameter Estimation, Naive Bayes
  • Asg 3 - Weight: 20% - Topics: Part II - Classification and Regression with Decision Trees and Ensemble Methods
  • Asg 4 - Weight: 20% - Topics: Part II and III - Deep Learning