Project

UFJF-MLTK

C++ machine learning toolkit (classification, regression, clustering, feature selection) co-authored at UFJF.

Archived Repository → Paper →

Overview

UFJF-MLTK is a C++ machine learning toolkit developed during my time as a researcher at UFJF. It covers classification, regression, clustering, feature selection, and ensemble methods, with a focus on a clean, header-organized C++17 design.

What’s in it

  • Classifiers: SVM (primal/dual variants), perceptron, KNN, ensembles.
  • Regressors: LMS, ridge.
  • Clustering: K-means, hierarchical.
  • Feature selection: wrapper methods built on top of any classifier.
  • Utilities: dataset abstractions, validation harness, metrics.

Status

Archived — used in coursework and the linked paper. Linked here as historical academic credit, not active engineering work.