Anthony Mitchell

Follow me on GitHub

Thoughts on Grad School

This document is intended to save some of my thoughts on grad school, and hopefully to help me answer the question “What did you learn in grad school?”.

When I graduated Wright State University with a BS in Computer Science, I had ZERO intention of going back for a masters degree. I had done some work with Wright State Research Institute and knew that academia wasn’t for me. I also really wanted to get into the industry and start my career. Within 6 months of working at Radiance, I knew I needed to get my masters degree because I was working with so many smart people. While education does not automatically equal intelligence, this was a good route for me to up my credentials and look better on proposals. I toyed with a couple of potential thesis topics, before deciding on the non-thesis route as a better option (Made this decision with several managers at Radiance, one a PHD in CS).

Fall 2022

  1. Intro Machine Learning - CS 6840 This course was taught by an adjunct lecturer that I really came to like, Sanhgui Han (No longer teaches, to my knowledge). She works full time at CFD Research, and several years later I got to teach her in an AFSIM training class! More of a specialized topic, but nothing I didn’t learn in an undergrad machine learning course. The final project proved to be fun, as I finally implemented and analyzed the MENACE algorithm (writeup someday?)!

The first gut punch of grad school happened in this class, when a fellow grad student used the chat forum to ask why their Python code was not working. pip install scikit learn was not working for them and they could not diagnose the error messages or follow any basic Google search for the FIRST STEP of every tutorial that tells you the correct import statement. Apparently scikit-learn was too much for them. I can look past spelling (I am a terrible speller and writer) and even some confusion if this class was the first time using Python, but I would expect better problem solving skills from a grad student. This was my first glimpse at how bad grad students could be. Roughly half of the 20 odd students in the class presented projects that were clearly half baked or not performed at all. She was very quick to point them out and I think that wore on her as a poor indication of the quality of the students. She apparently got her doctorate while on active duty in the middle east, dealing with internet blackouts as operations were being performed. A far cry from those not able to handle a Python import statement.

  1. Computability and Complexity - CS 7220 This was the very theory heavy class that temporarily made me rage but overall was good. It was very obviously taught by someone with pure theory knowledge, going over the strictest mathematical definition of Big O Notation. Covered FSA and ND-FSA as well as the titular complexity classes.

Spring 2023

  1. Algorithm Design and Analysis - CS 7200 Another theory heavy class by T.K. Prasad. It also is one of the CS core classes for a masters. It taught some interesting tricks and some patterns for the interview type scheduling/resource allocation questions you might get. While those specific scenarios might not pop up in day to day programming, it was good to have some more experience and tools for thinking about these sorts of problems.

  2. Soft Computing - CS 7840 Another special topics class taught by Sanhgui Han. The class was split into 3 topics; NLP, machine learning, and quantum computing. Being split up, it felt like more of a survey type class, and was fairly fun. The final presentations were again not very well done and she went off on several for not really doing the work.

Summer 2023

Working full time and taking 2 grad school courses was catching up to me, and with a shortened summer semester ahead of me, I decided to slow down and take one course at a time.

  1. Advanced Software Engineering - CS 7140 This one was taught by Eric Buck and was easier than the undergrad version I took with him. You listen to him ramble about whatever he decides to talk about, take some exams and pick a GitHub project to extend with a small team. Stupid easy. Another core course for the masters program, even though all these were things I knew from working with a team at work. Our team project was forking and slightly extending a Linux based Minecraft clone.

Fall 2023

  1. Advanced Programming Languages - CS 7100 A good theory course from T.K. Prasad. Implemented a simple language with LISP. His standard hard exams with low scores but good curve. Easy enough, though I was traveling a lot during this semester. I remember taking many notes in different airport lounges with Ted Masternak. Another core course where I noticed the trend of low exam scores with a high curve.

Spring 2024

  1. Information Retrieval - CS 7800 Classic T.K. Prasad class. 2 small python projects. One implemented a basic indexing and search application (academic, nothing grand like you think of when you hear “application”). Do some stats on it to show performance. Second one was extending some sklearn libs to create custom classifiers and doing the performance evaluations on them.

Fall 2024

  1. Fundamentals of Data Science - CS 7730 Pretty sure I was in the first offered section of this. It was stupid easy. 3 simple Python projects to plot basic data in an open ended effort. Based on my job, this could have been a really cool way to explore more unique ways to display data but it was a bit underwhelming. I remember thinking the professor was a bit of a pushover when it came to assignment due dates and assignment criteria, setting back the requirements and due dates multiple times.

Spring 2025

  1. Privacy Aware Computing - CS 7850 Professor Bin Wang. Very research and academic focused. Would have liked more involved examples to test the concepts presented so abstractly with random paper reviews and a final paper.

Fall 2025

  1. Human Computer Interface - CS 6900 This one caught my attention and I think was a good class to end on. The professor was relatively new to WSU and getting used to the student body there. I think that caused some issues with his expectations vs reality. The material was very usability focused, and had some neat projects to emphasize the topics. We created a new icon for a future technology device and had to think about how to convey the message with the simplest, most generic images. We created a selfie application that directed a user left and right, up and down as well as their face orientation to center them for a photo. This was on the topic of accessibility. We also did a study on color recognition for colorblind useability of different palates (Accessability and studies). The final app was a workout tracker than used pose estimation to track reps and sets of bicep curls (Python and CV). State tracking was a fun problem for these projects and I really enjoyed working on the utilities and overall structure of the program. This joy usually ended when my mandatory teammate would add her commits using AI and completely rewrite sections or misuse them. It was also really grating to get DMs that were obviously written by AI. Several teams shared the exact same UI for their colorblind study, leading to academic integrity issues and a slight breakdown of the professor during class. The final exam did not get good scores and received a similar disproving message from the professor, very similar to Sanghui Han’s. This was disappointing to see from my cohort and teammate specifically.

Written by a human.

For questions, bug reports, or complaints, please contact me!