Improving Models: Feature Engineering and Hyperparameters
Making Models Better
Feature engineering
Features are the pieces of input data a model uses. Feature engineering means choosing, transforming, or combining inputs so the model can learn more effectively.
Hyperparameters
Hyperparameters are settings chosen before or during training, such as learning rate, batch size, number of layers, and number of epochs. They are not learned directly from the data.
Common Hyperparameters
Improvement process
-
1
Start with a simple baseline model.
-
2
Check training and validation results.
-
3
Adjust one feature or hyperparameter at a time.
-
4
Compare results carefully.
-
5
Keep the best version and document what changed.
Which is a hyperparameter?
Learning rate is a training setting chosen by the user.
Correct answer: A learning rate
What is feature engineering?
Better features can improve model quality.
Correct answer: Choosing or transforming input data to help the model learn better.
Why change one thing at a time when improving a model?
Changing one factor at a time helps you understand what worked.
Correct answer: To make debugging easier and results clearer