Math With Python

A collection of math- and AI-related notebooks and scripts paralleling my studies in Linear Algebra, Statistics, Machine Learning, Deep Learning, LLM Model Mechanisms and AI Safety.

---

About the Repository

This repo reflects my homework/practice/projects as a self-taught developer learning mathematics and applied programming in an effort to work on AI Safety and Alignment.

It includes structured coursework-based exercises, hands-on problem solving, and will soon include independently developed exploratory projects.

---

Repo Structure

.
├── linearAlgebra/            # Exercises from Linear Algebra coursework
├── statsMachineLearning/     # Exercises from Stats & Machine Learning coursework
├── variousProjects/          # Independent math projects and experiments (Forthcoming!)
├── templates/                # Reusable notebook/script templates for notes
├── requirements.txt          # Project dependencies
└── LICENSE                   # MIT License

---

Study Sources

• Mike X. Cohen – Linear Algebra (Udemy)

• Mike X. Cohen – Statistics & Machine Learning (Udemy)

These courses emphasize a dual focus on mathematical understanding and Python/MATLAB implementation.

---

Highlights

From the most recent Jupyter notebooks:

• The t-test, et alia -- What I've learned about the t-test so far: assumptions, parameters, practical applications, constraints.

• Generalized Eigendecomposition: Two Methods -- Analyzing eigenvalues and eigenvectors of $A v = \lambda B v$ via both direct generalized eigendecomposition eig(A, B) and the equivalent standard eigenproblem form $B^{-1}A v = \lambda v$.

• Manual Eigendecomposition, 2x2 Case -- Generates a clean random 2x2 matrix. Computes eigenvalues and eigenvectors manually. Verifies the results via numpy.linalg.eig().

• Visualizing the Law of Large Numbers and the Central Limit Theorem -- Analyzes several distributions to demonstrate and visualize the law of large numbers and the central limit theorem.

• Complex Eigenvalue Art -- Harvests complex eigenvalues from random matrices. Makes something funky.

• Least Squares Modeling and Analysis -- Takes mock data, builds a Least Squares model, visualizes and evaluates the results.

---

Skills Practiced

• Math/ML/AI Domains:
  • Linear Algebra, Statistics, and Machine Learning fundamentals
  • LLM Model Mechanisms and Mechanistic Interpretability
• Python Development:
  • OOP and modular scripting
  • Jupyter Notebooks + LaTeX math rendering
  • NumPy, SymPy, Matplotlib, Seaborn, Plotly
  • Custom visualizations and exploratory analysis
• Web Development (parallel study):
  • Full-Stack JavaScript (Node.js, React, Express)
  • Flask (Python)
  • SQL/PostgreSQL databases

---

About Me

A self-taught, full-time student focused on Data Science, Mathematics, Software Development, and AI Safety Theory.

Coursework studied since 2022:

• Python for Math, Data Science, Application Development and Web Development.

• Linear Algebra, Statistics, Machine Learning, Calculus and Number Theory, implemented with Python.

• JavaScript (ESM/Express/React) for application development and full-stack web development.

Passionate about AI Alignment and Safety.

Open to internships, junior dev roles, and meaningful collaboration.

Trying to continually learn -- from bootcamps, online documentation/materials/books/audio/video, collaboration/conversation/apprenticeship, and building real things.

---

License

This project is licensed under the MIT License.

---

Andrew Blais – Boston, MA
GitHub: github.com/andrewblais
Website: wateronchair.com