Intro

This repostitory is a path through my graduate journey at Northwestern. I was initally accepted into the Analytics Management Certificate program, but wanted to go deeper into the program. There will be a seperate branch for the specific course, which I will share the projects/content in accordance to course/university policy.

Program overview: https://catalogs.northwestern.edu/sps/courses-az/graduate/msds/

Table of Contents

• About
• Demo
• Features
• Installation
• Interests
• Contributing
• Contact

About my Journey in Grad School

I was waived from MSDS 400, 401, and 476 due to adequate substitute courses. I enrolled in multiple algorithms courses, and earned an A at UIUC's CS 401. Additionally, my statistics background was also waived due to high performance in: Time Series Analysis, Applied Econometrics, and another statistical equivalance of Econometrics from computational side. For MSDS 476, I was waived due to performance in courses regarding: Performance Managements and Business Analytics Introduction (graduate level, simulation intro).

My first class was MSDS 460 (Summer 2024), which was extremely relevant. I recommend this be taken early on in the program, as it will lay foundation for deeper classes to come. It was an applied linear programming course, touching into heuristics. This should be a pre-requisite for MSDS 422 in my opinion - Dr. Kline is an amazing professor and makes the material interesting, and is very encouraging. The grading was generous.

• In between this course, I spoke with a professor in the PhD program, and realized that there will be a need to go deeper into the theoretical frameworks to truly understand and build original algorithms. I don't believe leet coding will be sufficient, so I am planning on enrolling in a bundle of theoretical courses, similiar to here:

  • Statistics Certificate is an additional path I am considering before the Thesis, to get a deeper understanding of the theory. - Sample book on introductory topics - Deeper Stochastic Book

    Before Machine Learning:

    • I have taken up to Calc III, Diff EQ, Linear Algebra, Linear Programming, Time Series Analysis - so those areas are covered.

    • Basic probability/statistics:

      • Foundational for probabilistic models and uncertainty estimation in machine learning.
      • Also need to develop understanding of foundations of Gradient Descent before the course.

      Before Deep Learning:

      • Numerical optimization (like SGD):

        • Essential for handling large datasets and training deep networks, supported by courses like Statistical Learning and Data Science.

      • Advanced calculus (like Jacobians and Hessians):

        • Learn more advanced mathematical concepts in Introduction to Theoretical Statistics.

      • Non-convex optimization: Many neural networks have non-convex loss functions, requiring deeper understanding of optimization methods.

      Before Simulation:

    • After ML and DL, focus on stochastic processes (like Markov processes),

      • numerical methods (e.g., finite differences), and optimization theory for handling constraints and probabilistic models (e.g., Lagrange multipliers).

My second and third course will be MSDS 420 and MSDS 485 (Fall 2024). 420 is a prerequisite for some other courses, and should give a solid introduction for accessing, organizing, and structuring data. I do have experience working with databases, so I decided to also enroll in MSDS 485 in this time period. I was able to peak ahead at MSDS 420 and felt comfortable with the balancing workloads here.

• Throughout and during this period, I am also reading some books related to devops, exploring functional programming, as well as working on a strong foundation in leetcode problems.

My fourth course will be MSDS 434 (Winter 2025). I reached out before the Fall to prepare a bit, and think of how to apply it to roles. I also wanted to allow ample time to prepare for MSDS 422 (Machine Learning), and will comment some other Math prep I did for that course.

My fifth course will be MSDS 422 (Spring 2025).

Demo of Classes

The above image shows the core courses necessary, in addition to MSDS 476. I waived three, so the other five are necessary. The seven electives I am thinking are:

• Analytics Systems Engineering (https://sps.northwestern.edu/masters/data-science/program-courses.php?course_id=4774)

• Artificial Intelligence and Deep Learning (https://sps.northwestern.edu/masters/data-science/program-courses.php?course_id=4783) pre-requisite and enabling to deeper into the two below.

  • Intelligent Systems and Robotics (https://sps.northwestern.edu/masters/data-science/program-courses.php?course_id=5012)
    • Pre-reading this textbook in advance Distributed Processing - Overlapping Topics
  • Applied Probability and Simulation Modeling (https://sps.northwestern.edu/masters/data-science/program-courses.php?course_id=4779)

• Data Pipelines and Stream Processing (https://sps.northwestern.edu/masters/data-science/program-courses.php?course_id=5289) OR Full Stack Data Engineering (https://sps.northwestern.edu/masters/data-science/program-courses.php?course_id=5054).

• Analytics Application Engineering Possible 6th course (https://sps.northwestern.edu/masters/data-science/program-courses.php?course_id=5011)

• The seventh course will be most related to understanding edge computing.

• Finally, the last step would be a Thesis Research project, which would be guided by Learning repository throughout the program of applied research and projects.

Interests

Refer to my Learning branch to see project applications.