NeuralNetworkClassifier

class NeuralNetworkClassifier(neural_network, loss='squared_error', one_hot=False, optimizer=None, warm_start=False, initial_point=None, callback=None)

Bases: TrainableModel, ClassifierMixin

Implements a basic quantum neural network classifier. Implements Scikit-Learn compatible methods for classification and extends ClassifierMixin. See Scikit-Learn for more details.

Parameters:

• neural_network (NeuralNetwork) -- An instance of an quantum neural network. If the neural network has a one-dimensional output, i.e., neural_network.output_shape=(1,), then it is expected to return values in [-1, +1] and it can only be used for binary classification. If the output is multi-dimensional, it is assumed that the result is a probability distribution, i.e., that the entries are non-negative and sum up to one. Then there are two options, either one-hot encoding or not. In case of one-hot encoding, each probability vector resulting a neural network is considered as one sample and the loss function is applied to the whole vector. Otherwise, each entry of the probability vector is considered as an individual sample and the loss function is applied to the index and weighted with the corresponding probability.

• loss (str | Loss) -- A target loss function to be used in training. Default is squared_error, i.e. L2 loss. Can be given either as a string for 'absolute_error' (i.e. L1 Loss), 'squared_error', 'cross_entropy', or as a loss function implementing the Loss interface.

• one_hot (bool) -- Determines in the case of a multi-dimensional result of the neural_network how to interpret the result. If True it is interpreted as a single one-hot-encoded sample (e.g. for 'CrossEntropy' loss function), and if False as a set of individual predictions with occurrence probabilities (the index would be the prediction and the value the corresponding frequency, e.g. for absolute/squared loss). In case of a one-dimensional categorical output, this option determines how to encode the target data (i.e. one-hot or integer encoding).

• optimizer (Optimizer | Minimizer | None) -- An instance of an optimizer or a callable to be used in training. Refer to Minimizer for more information on the callable protocol. When None defaults to SLSQP.

• warm_start (bool) -- Use weights from previous fit to start next fit.

• initial_point (np.ndarray) -- Initial point for the optimizer to start from.

• callback (Callable[[np.ndarray, float], None] | None) -- a reference to a user's callback function that has two parameters and returns None. The callback can access intermediate data during training. On each iteration an optimizer invokes the callback and passes current weights as an array and a computed value as a float of the objective function being optimized. This allows to track how well optimization / training process is going on.

Raises:

QiskitMachineLearningError -- unknown loss, invalid neural network

Attributes

callback	Return the callback.
fit_result	Returns a resulting object from the optimization procedure.
initial_point	Returns current initial point
loss	Returns the underlying neural network.
neural_network	Returns the underlying neural network.
num_classes	The number of classes found in the most recent fit.
optimizer	Returns an optimizer to be used in training.
warm_start	Returns the warm start flag.
weights	Returns trained weights as a numpy array.

Methods

fit(X, y)	Fit the model to data matrix X and target(s) y.
load(file_name)	Loads a model from the file.
predict(X)	Predict using the network specified to the model.
save(file_name)	Saves this model to the specified file.
score(X, y[, sample_weight])	Returns a score of this model given samples and true values for the samples.