Member-only story
What is Data preprocessing, Why Do We Need That❓
Preprocessing as the name suggests, is the processing of something before it is used or fed in some scenario. So Data preprocessing means, processing before the data is fed to the deep learning model. In every machine learning or other problem, where our main factor is data, Data preprocessing plays an important role that includes Data normalization, Data rescaling, Data reshaping, Data augmentation, removing outliers from data, or removing useless points from data, etc.
Data preprocessing techniques may vary with different data domains, i.e. preprocessing techniques for image data can be different from preprocessing techniques for textual data. But is mainly dependent on the use case.
Generally, Data is the first and one of the major factors regarding the performance and accuracy of machine learning models.
Why do we need that?
Sometimes we are dealing with a dataset that has not been used or tuned before with any machine learning model and It comes from the real world, where data and their insights are changing regularly. Here data scientists or data engineers will need to understand and find insights from data and make something that can integrate with real-world data. But ❓
Machine learning models do not know what is real data, they only know about machine language. So in that…
