Nonlinear Dimensionality Reduction Techniques: A Data Structure Preservation Approach

This book provides a comprehensive overview of tools for analyzing multidimensional and metric data, with a focus on visual exploration by human analysts.

It covers existing techniques such as tSNE and Isomap, as well as new unsupervised and supervised methods like ASKI, ClassNeRV, and ClassJSE.

The book also introduces MING, a novel approach for evaluating local map quality.

These methods are applied to the representation of expert-designed fault indicators for smart buildings, I-V curves for photovoltaic systems, and acoustic signals for Li-ion batteries.

With this book, you'll gain a deep understanding of how to analyze complex data sets and make informed decisions in the field of energy systems.