Review:
Z Score Scaling
overall review score: 4.5
⭐⭐⭐⭐⭐
score is between 0 and 5
Z-score scaling, also known as standardization, is a data preprocessing technique used to normalize features by subtracting the mean and dividing by the standard deviation. This transformation results in a distribution with a mean of zero and a standard deviation of one, which helps many machine learning algorithms converge faster and perform better.
Key Features
- Transforms features to have zero mean and unit variance
- Facilitates comparison across different scales
- Improves the performance of algorithms sensitive to feature scales (e.g., SVM, k-NN)
- Easy to implement and widely used in data preprocessing pipelines
- Does not distort the relationships between original data points
Pros
- Standardizes data, making it easier for models to learn
- Reduces bias caused by differing feature scales
- Widely supported with numerous libraries and tools
- Simple and computationally efficient
Cons
- Sensitive to outliers, which can skew the mean and standard deviation
- Assumes data is normally distributed, which may not always be true
- Not suitable for data with non-stationary distributions without additional techniques