An Effective Data Normalization Strategy for Academic Datasets using Log Values

Abstract

Data transformation is converting input data from one form to another form. Normalization is a standard data transformation techniques which can be used to transform data. There are many types of normalization techniques, like Min-Max normalization, Cube Root normalization, etc. In this paper, we implement a two-phase normalization process. In the first phase, the data is normalized using Cube Root normalization. Next, the normalized data is still more scaled by applying the logarithmic transformation. Finally, the performance of the transformed values is evaluated against standard metrics.