What if machine learning could be understood geometrically?
This book presents a unified geometric perspective on machine learning, statistics, and data science through the language of geometric algebra. From linear models and principal component analysis to neural networks, attention mechanisms, and time series systems, modern methods are reinterpreted as geometric transformations in n-dimensional spaces.
Rather than treating techniques as isolated tools, this book reveals the common structure underlying them: movement, orientation, and shape.
• Connects machine learning methods through geometry • Covers PCA, neural networks, attention, and time series • Includes PyTorch implementations • Bridges theory and real-world applications • Emphasizes intuition over formalism
