Notes by Lex Toumbourou

Loss function is a mathematical function that evaluates how well a machine learning model's predictions match actual outcomes.

I've written about a few different loss functions, that come up a lot in Machine Learning:

• Mean Absolute Error - one of the most basic loss functions. If you have a problem where you're trying to predict something continuous like the height of a person, the mean absolute error is simple the absolute different between your prediction and th action: abs(prediction - actual).
• Root Mean-Squared Error
• Binary Cross-Entropy Loss
• Categorical Cross-Entropy Loss