English
Languages
English
Bengali
French
Hindi
Italian
Japanese
Korean
Malayalam
Russian
Spanish
Tamil
Turkish
Vietnamese
Shortcuts

Source code for qiskit_machine_learning.neural_networks.two_layer_qnn

# This code is part of Qiskit.
#
# (C) Copyright IBM 2020, 2021.
#
# This code is licensed under the Apache License, Version 2.0. You may
# obtain a copy of this license in the LICENSE.txt file in the root directory
# of this source tree or at http://www.apache.org/licenses/LICENSE-2.0.
#
# Any modifications or derivative works of this code must retain this
# copyright notice, and modified files need to carry a notice indicating
# that they have been altered from the originals.

"""A Two Layer Neural Network consisting of a first parametrized circuit representing a feature map
to map the input data to a quantum states and a second one representing a ansatz that can
be trained to solve a particular tasks."""
from typing import Optional, Union

from qiskit import QuantumCircuit
from qiskit.circuit.library import RealAmplitudes, ZZFeatureMap
from qiskit.opflow import PauliSumOp, StateFn, OperatorBase, ExpectationBase
from qiskit.providers import BaseBackend, Backend
from qiskit.utils import QuantumInstance

from .opflow_qnn import OpflowQNN
from ..exceptions import QiskitMachineLearningError


[docs]class TwoLayerQNN(OpflowQNN): """Two Layer Quantum Neural Network consisting of a feature map, a ansatz, and an observable. """ def __init__( self, num_qubits: int = None, feature_map: QuantumCircuit = None, ansatz: QuantumCircuit = None, observable: Optional[OperatorBase] = None, exp_val: Optional[ExpectationBase] = None, quantum_instance: Optional[Union[QuantumInstance, BaseBackend, Backend]] = None, input_gradients: bool = False, ): r""" Args: num_qubits: The number of qubits to represent the network, if None and neither the feature_map not the ansatz are given, raise exception. feature_map: The (parametrized) circuit to be used as feature map. If None is given, the `ZZFeatureMap` is used. ansatz: The (parametrized) circuit to be used as ansatz. If None is given, the `RealAmplitudes` circuit is used. observable: observable to be measured to determine the output of the network. If None is given, the `Z^{\otimes num_qubits}` observable is used. input_gradients: Determines whether to compute gradients with respect to input data. Note that this parameter is ``False`` by default, and must be explicitly set to ``True`` for a proper gradient computation when using ``TorchConnector``. Raises: QiskitMachineLearningError: In case of inconsistent num_qubits, feature_map, ansatz. """ # check num_qubits, feature_map, and ansatz if num_qubits is None and feature_map is None and ansatz is None: raise QiskitMachineLearningError( "Need at least one of num_qubits, feature_map, or ansatz!" ) num_qubits_: int = None feature_map_: QuantumCircuit = None ansatz_: QuantumCircuit = None if num_qubits: num_qubits_ = num_qubits if feature_map: if feature_map.num_qubits != num_qubits: raise QiskitMachineLearningError("Incompatible num_qubits and feature_map!") feature_map_ = feature_map else: feature_map_ = ZZFeatureMap(num_qubits) if ansatz: if ansatz.num_qubits != num_qubits: raise QiskitMachineLearningError("Incompatible num_qubits and ansatz!") ansatz_ = ansatz else: ansatz_ = RealAmplitudes(num_qubits) else: if feature_map and ansatz: if feature_map.num_qubits != ansatz.num_qubits: raise QiskitMachineLearningError("Incompatible feature_map and ansatz!") feature_map_ = feature_map ansatz_ = ansatz num_qubits_ = feature_map.num_qubits elif feature_map: num_qubits_ = feature_map.num_qubits feature_map_ = feature_map ansatz_ = RealAmplitudes(num_qubits_) elif ansatz: num_qubits_ = ansatz.num_qubits ansatz_ = ansatz feature_map_ = ZZFeatureMap(num_qubits_) self._feature_map = feature_map_ input_params = list(self._feature_map.parameters) self._ansatz = ansatz_ weight_params = list(self._ansatz.parameters) # construct circuit self._circuit = QuantumCircuit(num_qubits_) self._circuit.append(self._feature_map, range(num_qubits_)) self._circuit.append(self._ansatz, range(num_qubits_)) # construct observable self.observable = ( observable if observable else PauliSumOp.from_list([("Z" * num_qubits_, 1)]) ) # combine all to operator operator = ~StateFn(self.observable) @ StateFn(self._circuit) super().__init__( operator, input_params, weight_params, quantum_instance=quantum_instance, input_gradients=input_gradients, ) @property def feature_map(self) -> QuantumCircuit: """Returns the used feature map.""" return self._feature_map @property def ansatz(self) -> QuantumCircuit: """Returns the used ansatz.""" return self._ansatz @property def circuit(self) -> QuantumCircuit: """Returns the underlying quantum circuit.""" return self._circuit @property def num_qubits(self) -> int: """Returns the number of qubits used by ansatz and feature map.""" return self._circuit.num_qubits

© Copyright 2018, 2021, Qiskit Machine Learning Development Team.

Built with Sphinx using a theme provided by Read the Docs.