A complete data science pipeline built in C++ using Eigen, with Python/matplotlib visualization.
- Loads forensics file-integrity data from CSV
- Computes descriptive statistics (mean, variance, std dev)
- Trains a logistic regression classifier from scratch
- Predicts hash mismatch (suspicious vs clean files)
- Outputs results to CSV for Python visualization
- C++17 + Eigen (numerical computation)
- CMake + vcpkg (build system)
- Python + matplotlib + pandas (visualization)
Achieved 100% accuracy on 30-sample forensics dataset after 5000 epochs of gradient descent.
Most data science is Python. This project deliberately implements the full pipeline in C++ to understand what libraries like NumPy and sklearn actually do underneath.