Data plays a vital role in research, providing valuable insights that guide decision-making. However, raw data is often messy and unstructured, making it difficult to use effectively. That’s why data preprocessing is an essential step, it helps organize the data, ensuring consistency and improving the performance of machine learning models, even though it can be time-consuming. This project explores how selected features can improve the accuracy of machine learning models. We are using an existing structured dataset from Amazon review and focused on identifying features in the data that can improve predictions. In addition, we will train machine learning models or experiment with automated machine learning (AutoML) tools. Our goal is to build models that not only perform well during training but also make accurate predictions on new data. This study emphasizes the value of selecting the right features to create reliable and effective machine learning models.
