PolicyGradient class

REINFORCE-style Policy Gradient with optional baseline and normalization

Constructors

PolicyGradient({required int nActions, required List<int> policyLayers, int? seed, bool useBaseline = false})

Properties

baseline ↔ double

getter/setter pair

hashCode → int

The hash code for this object.

no setter, inherited

nActions → int

final

policy → ANN

final

runtimeType → Type

A representation of the runtime type of the object.

no setter, inherited

useBaseline → bool

final

Methods

noSuchMethod(Invocation invocation) → dynamic

Invoked when a nonexistent method or property is accessed.

inherited

selectAction(List<double> state) → int

toJson() → String

toMap() → Map<String, dynamic>

toString() → String

A string representation of this object.

inherited

updateFromEpisode(List<List<double>> states, List<int> actions, List<double> returns, {bool normalize = true}) → void

Update from a single episode represented as lists of states, actions and returns. Supports optional baseline subtraction and advantage normalization.

Operators

operator ==(Object other) → bool

The equality operator.

inherited

Static Methods

fromJson(String s, {int? seed}) → PolicyGradient

fromMap(Map<String, dynamic> m, {int? seed}) → PolicyGradient