Example: Difference between revisions

In [[machine learning]], [[example]]s or [[training data]] are [[input]] data and the corresponding desired [[output]]. A single [[example]] is the values of one row of [[features]] and possibly a label. Examples in [[supervised learning]] fall into two general categories: labeled examples and unlabeled examples. Labeled examples comprise one or more [[features]] and a [[label]]. Labeled examples are used during [[training]]. On the other hand, unlabeled examples consist of features but no label. Unlabeled examples can be utilized during training and [[inference]].

An example in machine learning refers to a pair of input and output values used to train a model. The input value is made up of [[features]] or attributes that describe an object or phenomenon, while the output value serves as its [[label]] or [[class]]. For instance, spam detection systems typically take email messages (input) as their label (output), which could either be "spam" or "not spam." Similarly, image recognition systems take pictures as inputs and assign them labels describing what the image depicts. Suppose we collected a dataset of 2000 apartments with their features and prices. The features and the price of a single apartment would be an example.