MATH 490 Capstone Notes and Updates

Last Updated: 2024-10-01

Date: 2024-09-27

  • Met with Dr. McGuirk, talked more about finalizing talks.
    • Looking into conferences and symposiums
    • Honestly after looking through this stuff, I looked through NYU, Columbia, and just locally, theres not many to go to in a 3 month span. All of them require a good bit of money.
    • I think I will set up an event at the school, and invite faculty and students to come purely for lecturing practice. I will flesh out this talk as I go, but I'm leaning towards talking about a PCA analysis, and using League gameplay data to do it. Depending on how far along in my Neural Network progress I get, maybe a brief prototype
  • What am I working on?
    • Book that I'm following: Neural Networks Chapter 1
    • What did I work on for the week: Learning about GANS (Generative Adversarial Networks) The original GAN paper
    • Working on a proof of concept having a neural network to recognize handwritten digits.
    • Also working on writing up Paolo Gasti's machine learning algorithm linked below.
  • Homework
    • Finish reading through Paolo Gasti's Paper and develop some prototype of the algorithm
    • Reading about alphas and degrading, Black-Scholz, Stone-Weirstrauss theorem, and the Alignment Problem.
    • Develop prototype with 4-output neural network to recognize handwritten digits
    • Problems about distributions and probability theory
    • Make powerpoints for a talk, and flesh out my progress in short talks to give Dr. Katsnelson and Dr. Roy

Date: 2024-09-20

  • Met with Dr. Mcguirk
    • Discussed project ideas
      • Machine Learning Algorithm for Finance based on word count events in the news
      • Draft Analysis for upcoming League of Legends World Championship
      • Auction House (Stock Market) analysis for World of Warcraft, finding profit margins
      • Paolo Gasti's ML paper
      • (Proof of Concept) Yahoo Webscraper
    • Discussed project requirements
      • Strong Github portfolio
      • 3 talks need to be given: One at a conference, one in a room with Dr. Roy and faculty, and one at a seminar perhaps. (This isn't fully fleshed out, but we're working on it)
      • I need to demonstrate proof of concept each week (If you're reading this Dr. Katsnelson, hello!)
  • Set up a foundation for probability theory in Machine Learning
    • Discussed sample spaces, events, and probability distributions
    • Countability of infinite spaces