I love coding, mathematics and solving complex world challenges by applying technology to medicine, bioscience and intelligence analysis.

I’ll publish some of my personal work here.

Dimensionality Reduction Exploration

This project explores different methods for dimensionality reduction using the Wine dataset.
The main goal is to compare Principal Component Analysis (PCA) with Linear Autoencoders (LAE) and understand their similarities and differences in terms of latent space representation.

📂 Repository

👉 GitHub Repository

📑 Notebooks

Principal Component Analysis (PCA)
Implements PCA from scratch: eigen decomposition of covariance matrix, explained variance, and projection into reduced dimensions.

Linear Autoencoder (LAE)
Builds and trains a linear autoencoder in PyTorch, compares latent representations to PCA, and explores rotation equivalence.