The Effects of Data Standardization and Normalization Techniques in Click Through Rate Prediction

Abstract

Accurate ranking is critical for the user experience as well as applications such as information retrieval, recommender systems, and decision-making. To transform data into a common scale or distribution, standardization and normalization techniques are used. The purpose of this paper is to look into the effects of various data standardization and normalization techniques on ranking performance in order to improve performance or reduce computational complexity. It examines methods such as z-score standardization, min-max scaling, and robust scaling in existing literature and experimental studies. The paper assesses their impact on various ranking algorithms and models using benchmark datasets and discusses the benefits, limitations, and trade-offs associated with each technique, taking into account factors such as data distribution characteristics, outliers, and interpretability. The results can aid in the selection of the best normalization and standardization techniques for ranking tasks, particularly in recommender systems.