Beyond Asymptotics: Real-Time Performance of Basic Operations (Preliminary Results)

This paper presents preliminary findings on the real-time performance of fundamental mathematical operations, moving beyond traditional asymptotic analysis. While asymptotic analysis offers valuable insights into algorithms' theoretical behavior, it is equally important to consider real-world applications' practical performance and resource constraints. Our work empirically measures and analyzes the execution time of basic operations (addition, subtraction, multiplication, division, etc.) across various data types and sizes, along with a basic Artificial Neural Network (ANN). The results emphasize the effect of size on the algorithm's overall performance, keeping in mind the complication of the operation itself. Our results focus on studying real-time performance beyond the theoretical analysis offered by asymptotic analysis are considered the building block for further research. Moreover, it lays the groundwork for a deeper understanding of real-time performance for scalable algorithms for real-time big data analytics where efficient execution of basic operations is essential. Also, applying adaptive optimization based on machine learning techniques for optimal performance. Using machine learning to predict the real-time cost of basic operations is an innovative approach.