This paper aims to introduce various types of useful operations on fuzzy soft sets (FSSs), such as Einstein sum, Einstein product, algebraic sum, algebraic product, bounded product, bounded sum, and the basic properties of these new operations have been investigated. In addition, on the basis of our newly defined operations, we implement a new machine learning algorithm to solve FSS based real-life decision-making problems (DMPs), and using one real-life example we have demonstrated the viability of our proposed approach in real-world applications.