3 Training and Testing Models Free Online FlashCards

What happens during model training?

Training uses examples to adjust a model’s internal values so its outputs better match known answers.

What are model parameters?

Parameters are internal values that the model adjusts during training to improve its outputs.

What does a supervised learning example pair?

In supervised learning, each example pairs input features with a known target, such as a home’s characteristics paired with its sale price.

What does a loss function measure?

A loss function measures how far a model’s predictions are from the known targets.

What is the role of an optimization method in training?

An optimization method adjusts model parameters to reduce the loss.

How can training proceed across examples?

Training often repeats over examples or batches, making multiple passes through the data until a chosen stopping rule is met.

How do learning and inference differ?

Learning is the training phase that determines or adjusts parameters; inference applies the learned parameters to produce an output.

Does inference ordinarily change model parameters?

Inference does not ordinarily update the model’s parameters.

When does a new home become a training example?

A new home’s size is an inference input, not a training example unless its known sale price is later used in training.

What is a model prediction?

A prediction is the output a model produces for an input.

How can regression and classification predictions differ?

A regression model may predict a numerical value; a classification model may produce a class, a score, or both.

How can an email classifier turn a score into a label?

An email classifier can assign a spam score, then apply a decision rule to label the email “spam” or “not spam.”