Computational Trade-offs Between Traditional Algorithms and Machine Learning Models: A Time Complexity Perspective

Abstract: A fundamental distinction exists between the time complexity of traditional algorithms and machine learning (ML) algorithms. Traditional algorithms are used to solve specific problems by following a set of instructions. Machine learning algorithms are created to extract knowledge from data and apply what they have learned to new, unobserved data. In this research work, we find that there are dissimilar time complexity features between traditional algorithms and machine learning algorithms and therefore we suggest that their evaluation criteria should differ from each other while we perform an efficiency analysis of traditional algorithms and machine learning algorithms. We distinguish this research work from prior related research works by examining the relative performance of machine learning models and traditional algorithms in terms of training and inference time. Keywords: Time complexity, machine learning, algorithms, efficiency analysis.