Local Explanation Method for Dimensionality Reduction Algorithms

• Grant number: 24/10791-0
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: August 01, 2024
• End date: June 30, 2025
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: José Alberto Cuminato
• Grantee: Lucas Greff Meneses
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Associated research grant: 13/07375-0 - CeMEAI - Center for Mathematical Sciences Applied to Industry, AP.CEPID

Abstract

Dimensionality reduction algorithms, such as UMAP (Uniform Manifold Approximation and Projection) and t-SNE (t-distributed Stochastic Neighbor Embedding), are widely used to simplify complex datasets by reducing the number of variables while preserving essential information as much as possible. This allows for various tasks, such as data visualization or training machine learning models.However, the transformative nature of these algorithms can make the results difficult to interpret. The objective of this project is to develop a local and algorithm-independent explanation method that can be applied to any dimensionality reduction technique, allowing for an understanding of how the original variables influence the reduced projection, thus aiding in informed decision-making across various applications.