Investigating the impact of data scaling on the k-nearest neighbor algorithm

This study investigates the impact of data scaling techniques on performance k-nearest neighbor (KNN) algorithm using ten different datasets from various domains. Three commonly used techniques, min-max normalization, Z-score, and decimal scaling, are evaluated based KNN algorithm's in terms accuracy, precision, recall, F1-score, runtime, memory usage. The aims to provide insights into applicability effectiveness contexts, aid design implementation machine learning systems, help identify strengths weaknesses each technique their suitability for specific types data. results show that significantly affects algorithm, choice method can have significant implications practical applications. Moreover, three varies across datasets, suggesting should be made characteristics Overall, this provides a comprehensive analysis practitioners researchers community make informed decisions when designing implementing systems.