mlpy.mdp.IMDPModel

class mlpy.mdp.IMDPModel(proba_calc_method=None)[source]

Bases: mlpy.modules.UniqueModule

The Markov decision process interface.

All Markov decision process (MDP) models are derived from the base class. The base class maintains an initial probability distribution from which the initial state can be sampled.

Parameters:

proba_calc_method : str

The method used to calculate the probability distribution for the initial state. Defaults to DefaultProbaCalcMethod.

Attributes

mid The module’s unique identifier.

Methods

fit(obs, actions, **kwargs) Fit the model to the observations and actions of the trajectory.
load(filename) Load the state of the module from file.
predict_proba(state, action) Predict the probability distribution.
sample([state, action]) Sample from the probability distribution.
save(filename) Save the current state of the module to file.
update(experience) Update the model with the agent’s experience.