Welcome to the definitive, first-principles repository tracking my progression through advanced Exploratory Data Analysis (EDA), feature engineering pipelines, numerical taxonomies, and machine learning models applied to complex datasets and high-frequency quantitative market dynamics.

This repository isn't just a collection of code snippets—it's a production-grade Obsidian Vault designed to bridge abstract mathematical theory with hard, real-world execution.

The vault is systematically divided into 5 foundational pillars, protecting model loss functions from noise, stabilizing matrix algebra, and maximizing predictive generalization.

📂 ml-notes/
├── 📂 EDA/                 # Visual distribution analysis, grid sweeps, and data ingestion
├── 📂 Feature-Engineering/ # Structural transformations, log scaling, and temporal spans
├── 📂 Feature-Selection/   # Regularized optimization (Lasso/Ridge) and VIF multi-collinearity checks
├── 📂 Statistics/          # Hypothesis testing, statistical derivations, and metrics intuition
└── 📂 Machine-Learning/    # System simulations, custom estimators, and portfolio architectures

1. Clone this repository directly onto your local system:

   git clone [https://github.com/shipitdev/ml-notes.git](https://github.com/shipitdev/ml-notes.git)

2. Open your Obsidian desktop application.
3. Select Open folder as vault and point it to the cloned ml-notes root directory.
4. Toggle the Graph View to dynamically inspect the cross-linked mathematical dependencies and tutorial paths.

This architecture framework repository is open-source software licensed under the MIT License