Dimensionality Reduction with Unsupervised Nearest Neighbors

Focusing on a novel dimensionality reduction technique, the book explores unsupervised nearest neighbors (UNN) as a method for enhancing classification and regression tasks. It begins with foundational machine learning concepts and a practical application in the energy sector. The text systematically develops various UNN models, addressing challenges like incomplete data and noise, while comparing different optimization strategies, including evolutionary and swarm-based methods. Richly illustrated with color figures, it presents experimental results that showcase UNN's effectiveness in both synthetic and real-world datasets.