👉 Regularisation is a technique used in machine learning and statistics to make predictions or models more stable. It's often applied when the model has parameters that are too large, which can cause overfitting. By normalising these parameters, we can reduce their size without changing the accuracy of our model. This method helps prevent overfitting and improve generalisation by making predictions on new data less sensitive to the initial training set's features.
