Classificação de séries temporais por meio de diferentes conjuntos de dados combinados com três classificadores distintos

Abstract

A time series is a sequence of data observed sequentially over time, with interde pendence between consecutive observations. This data is crucial in areas like finance, marketing, and economics, where it’s analyzed and modeled to predict future events, identify complex patterns, and extract valuable insights. In this study, the KNN (K Nearest Neighbors) method was combined with others like SVM (Support Vector Machine) and Random Forest to explore significant performance differences among them. Eight diverse datasets were selected and subjected to various normalization techniques. The analysis focused on understanding how data metrics and normalization impact method performance. Variations and trends were observed in each method and normal ization technique, aiming to provide insights into their effectiveness across different time series analysis contexts.