Skip to content

Commit

Permalink
Reword "models are data structures ..." (tensorflow#153)
Browse files Browse the repository at this point in the history
* Reword "models are data structures ..."

@lattner pointed out that this could be made more clear. Updated to

> Machine learning models are concepturally functions with internal state called "parameters". In code, models are often represented as a data structure that stores parameters as mutable properties and has an "apply" method.

* Update docs/ParameterOptimization.md

Co-Authored-By: rxwei <[email protected]>
  • Loading branch information
rxwei authored Mar 19, 2019
1 parent f857c7f commit 0e76995
Showing 1 changed file with 1 addition and 1 deletion.
2 changes: 1 addition & 1 deletion docs/ParameterOptimization.md
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,7 @@ The concept of parameter optimization is crucial for machine learning algorithms

### Parameters and optimization

Machine learning models are data structures with mutable properties called parameters. Machine learning optimizers "train" models by applying an algorithm (e.g. stochastic gradient descent) to update the parameters of a model.
Machine learning models are conceptually functions with internal state called "parameters". In code, models are often represented as data structures that store parameters as mutable properties and have an "apply" method. Machine learning optimizers "train" models by applying an algorithm (e.g. stochastic gradient descent) to update the parameters of a model.

In Swift, this might look like:

Expand Down

0 comments on commit 0e76995

Please sign in to comment.