What is data normalization, and why is it important in machine learning and data preprocessing? How does normalization differ from standardization? What are the common techniques used for normalizing data? How does normalization impact the performance of machine learning models? In what scenarios is data normalization especially useful?