A Comparative Study of Z-Score and Min-Max Normalization for Rainfall Classification in Pekanbaru

Abstract

Data preprocessing plays a crucial role in enhancing the performance of machine learning algorithms for classification tasks. Among the essential preprocessing stages is data normalization, which aims to standardize data into a comparable range of values. This study focuses on normalizing rainfall data in Pekanbaru from 2019 to 2023. The objective is to compare various data normalization techniques, including Min-Max Normalization and Z-Score Normalization. The comparison of these particular strategies is justified because they are widely applied and have different approaches. Min-max normalization is an easy-to-implement technique that makes the data sensitive to outliers by scaling it to a specific range, often from 0 to 1. However, Z-Score Normalization, sometimes referred to as Standardization, standardizes the data by dividing by the standard deviation and subtracting the mean, maintaining the shape of the distribution and making it resistant to outliers. The findings demonstrate that applying normalization techniques effectively enhances classification performance compared to using unnormalized data. Specifically, the optimal classification performance is achieved through Z-Score Normalization, yielding accuracy, sensitivity, and specificity rates of 74.59%, 82.48%, and 63.92%, respectively.