Which of the following statements accurately describe loss functions in machine learning? Select all that apply.

Loss functions measure how far a model’s predictions are from the true values, and the learning algorithm tries to make this value as small as possible. They are expressed as a formula that takes predictions and actual labels and returns a single number, often called the cost or error. By taking the derivative of this number with respect to the model’s parameters, we can use gradient descent to adjust the model. For example, in a simple linear regression the mean squared error loss is the average of squared differences between predicted and actual y values, and its derivative tells us how to shift the line to reduce the error. Thus, a loss function is a mathematical tool that quantifies prediction error and guides the optimization of the model.