Trait rurel::strategy::learn::LearningStrategy

source · 
pub trait LearningStrategy<S: State> {
    // Required method
    fn value(
        &self,
        new_action_values: &Option<&HashMap<S::A, f64>>,
        current_value: &Option<&f64>,
        received_reward: f64
    ) -> f64;
}
Expand description

A learning strategy can calculate a learned value for the action which was taken from the values for the actions in the new state (new_action_values), the current value (current_value), and the reward that was received after taking the action.

Required Methods§

source

fn value( &self, new_action_values: &Option<&HashMap<S::A, f64>>, current_value: &Option<&f64>, received_reward: f64 ) -> f64

Calculates a learned value for the action which was taken from the values for the actions in the new state (new_action_values), the current value (current_value), and the reward that was received after taking the action.

Implementors§