AbstractVedic Mathematics (VM) is an ancient system of mathematical techniques derived from Indian scriptures that offers a unique framework that emphasizes simplicity, mental agility, and pattern-based problem-solving. While traditionally applied to arithmetic and algebra, the core principles of VM show promise for modern computational domains such as machine learning (ML). In this paper, we have discussed how Vedic Sutras, namely Urdhva-Tiryakbhyam, Anurupyena, and Sankalana-Vyavakalanabhyam, can be adapted to optimize operations in ML. We examine their application in matrix multiplication, feature scaling, and gradient estimation and highlight results that demonstrate their efficacy. The integration of these ancient principles into cutting-edge ML models, such as Convolutional Neural Networks (CNNs), opens new directions for performance optimization and algorithmic innovation.
Kabra et al. (Wed,) studied this question.