Standardization and Normalization are not mandatory approaches and can be avoided in some cases, but some models like neural networks, SVM, etc. require the data to either be standardized or normalized. You can try running the model with and without standardization or normalization and evaluate the performance. If performance improves with standardization or normalization then keep it, if performance declines then remove it
Articles in this section
- Mounting Google Drive to Google Collaboratory Notebooks
- Binning data in Pandas
- Reading Python Notebook in Local Machine
- Pearson Correlation and Cosine Similarity
- How to display a specific number of rows in a pandas dataframe?
- How to display a particular number of columns in a pandas dataframe?
- What is a lambda function?
- Inclusion of number of columns in analysis when data contains large number of features
- Error in renaming a column
- Reading a .docx file in python