Minmaxskalning

Minmaxskalning, also known as Min-Max scaling or normalization, is a data preprocessing technique used in machine learning. It transforms features to a fixed range, typically between 0 and 1. This is achieved by subtracting the minimum value of the feature from each data point and then dividing by the range of the feature (maximum value minus minimum value). The formula for Min-Max scaling is: X_scaled = (X - X_min) / (X_max - X_min).

This method is particularly useful when algorithms are sensitive to the scale of input features, such as support vector machines (SVMs), k-nearest neighbors (KNN), and neural networks. By bringing all features to a similar scale, minmaxskalning prevents features with larger values from dominating those with smaller values, leading to more balanced model performance.

However, minmaxskalning is sensitive to outliers. An extreme value can drastically affect the minimum or maximum, thereby compressing the rest of the data into a very small range. If outliers are present and are not handled beforehand, they can lead to a suboptimal scaling. The output range can be adjusted if a different range, other than [0, 1], is desired. For example, scaling to [-1, 1] can be achieved with a slight modification of the formula.