negmas.models.future

Modeling self’s future prospects in the negotiation.

class negmas.models.future.FutureUtilityRegressor(regressor_factory=sklearn.gaussian_process.GaussianProcessRegressor, **kwargs)

Bases: object

Represents a regressor for own-utility for the future of the negotiation.

Remarks:

• We assume that the negotiation goes from time 0 to 1 (relative_time).

fit(times, utils) → FutureUtilityRegressor

Fit.

Parameters:

• times – Times.

• utils – Utils.

Returns:

The result.

Return type:

FutureUtilityRegressor

predict_time(utils) → ndarray

Predict time.

Parameters:

utils – Utils.

Returns:

The result.

Return type:

np.ndarray

predict_time_prob(utils, return_cov=False) → ndarray

Predict time prob.

Parameters:

• utils – Utils.

• return_cov – Return cov.

Returns:

The result.

Return type:

np.ndarray

predict_utility(times) → ndarray

Predict utility.

Parameters:

times – Times.

Returns:

The result.

Return type:

np.ndarray

predict_utility_prob(times, return_cov=False) → ndarray

Predict utility prob.

Parameters:

• times – Times.

• return_cov – Return cov.

Returns:

The result.

Return type:

np.ndarray