Faculty Collaborator: Andras Zsom
About:
Kameel collaborated with Professor Andras Zsom to develop the course DATA 2060: Machine Learning: From Theory to Algorithms. Building on a similar course, CSCI 1420: Machine Learning, Kameel created problem sets for students that involved recreating classical machine learning algorithms from scratch using Python. To facilitate this, he utilized Jupyter Notebooks, making the assignments easily accessible for students.
Agenda:
Project Goals and Expectations
Timeline and Challenges
Progress and Personal Growth
Questions
Project Goals and Expectations
Main Project Goals:
Work with Professor Andras Zsom to develop the course DATA 2060 - Machine Learning: From Theory to Algorithms.
Create Jupyter Notebook assignments that implement classical machine learning algorithms from scratch.
Each assignment includes a written portion, coding portion, and occasionally an ethics section.
Expectations:
Weekly meetings with Professor Zsom to discuss difficulties and progress.
Create Jupyter Notebooks, solutions, and unit tests for each of the 13 planned assignments for the next semester.
Prepare to hand over progress to TAs for the next semester to finish any outstanding work.
Timeline and Challenges
Project Timeline:
Assignments covered topics starting with Linear Regression and finishing with Neural Networks.
Each assignment was created using Markdown and Python within a Jupyter Notebook and stored in a GitHub repository for Professor Zsom and TAs to access next semester.
Workflow: Create Assignment → Create Solutions → Check over Assignment → Repeat.
Challenges:
Unfamiliarity with GitHub:
Attended Brown’s DSCoV GitHub Workshop.
Creating coding solutions for assignments:
Collaborated with Professor Zsom to address solutions and bugs.
Reviewed content from CSCI 1420 - Machine Learning.
Consulted future TAs of DATA 2060 who had taken CSCI 1420 for additional help.
Progress and Personal Growth:
Progress and Achievements:
Out of 13 planned assignments:
Completed 10 assignments.
Finished solutions and unit tests for 5 assignments.
Gained significant knowledge about the math behind machine learning algorithms and Python. Initially only familiar with basic ML algorithms and predominantly used R over Python.
Personal Growth:
Improved Technical Data Science Skills:
Linear Algebra, Statistics, and Mathematical Analysis.
Programming in Python.
Enhanced ability to communicate with peers and professors about difficulties to arrive at solutions.
Learned to provide timely and concise updates about progress, next steps, and timeline changes.