# Deep Learning Vocabulary*Glossary List*

- Inputs, outputs
* Inputs: What is fed into the black box function or model. Outputs: what is returned by the function/model.*
- Features, labels
* Features are the data attributes, raw data. Labels are the classes that the model should predict from each row of the features data.*
*Features, labels, models, functions* if we define a simple functional model, and give it a name f. f is a function, conceptually it is a model, which takes input X like this f(X),
the function maps X to y like this y=f(X). Big X is the feature matrix. Small y is the target vector we are trying to predict. Target y contain classes / labels in categorical prediction, versus values in regression, numeric predictions.
If you hear that functions are deterministic, that means for a specific value of X, it will always map it to the same y. We can get predictable result (same y) using the same X and function which maps X to y.
*Features :* The columns are attributes, feature names. Each of the row is a data record. For example, if we collect data from 100 people, there will be 100 rows of data in the feature matrix.