In L2 Regularization we find-

L2 Regularization, also known as Ridge Regression, is a technique used in machine learning to prevent overfitting. Here's how it works, step by step:

1. Overfitting: This is a common problem in machine learning where a model performs well on the training data but poorly on unseen data (like test data). This happens because the model learns the noise in the training data, making it complex and less generalizable.

2. Regularization: To prevent overfitting, we use a technique called regularization. Regularization adds a penalty term to the loss function, which discourages the learning algorithm from assigning too much importance to any individual feature, thus making the model simpler and more generalizable.

3. L2 Regularization: In L2 regularization, the penalty term is the square of the magnitude of the coefficients of the features. This is added to the loss function. The formula for the loss function with L2 regularization is: Loss function = Original loss function + λ*(sum of squares of coefficients). Here, λ is the regularization parameter. It determines how much importance we want to give to the penalty term. Higher λ will result in smaller coefficients, thus simpler model.

4. Training with L2 Regularization: When we train the model with L2 regularization, the model tries to minimize the loss function, which now includes the penalty term. This results in a model that not only fits the data but also keeps the coefficients of features as small as possible.

5. Result: The result is a model that is less likely to overfit, as it is discouraged from assigning too much importance to any individual feature. This makes the model more robust and better at generalizing to unseen data.