Quantum Circuit Data Encoding via Scaled Rotation Angles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current quantum machine learning methods face inefficiencies in encoding data within quantum circuits, limiting the expressivity and scalability of parametrized quantum circuits, particularly in utilizing the Hilbert space effectively for training procedures.
Innovation Solution
The proposed method involves applying encoding quantum gates that rotate qubits by rotation angles proportional to input features and scaling factors, which are powers of two, followed by variational quantum gates, allowing for measurement value determination and quantum circuit adjustments to enhance data encoding and output generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional encoding methods are used, then the quantum circuit can process data, but the expressivity and scalability are limited due to inefficient Hilbert space utilization
Solution Approach 1:
The patent applies parameter changes by introducing scaling factors (powers of two) to the rotation angles of encoding quantum gates. This transforms the encoding process from a linear to an exponential representation, where each qubit encodes multiple basis functions simultaneously. The rotation angle becomes proportional to input_feature × scaling_factor, enabling the quantum circuit to represent exponentially more basis functions with the same number of qubits, thereby resolving the contradiction between expressivity and device complexity.
2Adaptability or versatility
If additional qubits or encoding repetitions are added to increase expressivity, then more basis functions can be represented, but the device complexity and resource requirements increase
Solution Approach 1:
The patent inverts the conventional approach by not adding more qubits to increase expressivity, but rather by changing how existing qubits are utilized. Instead of each qubit representing one basis function, the scaled rotation angles enable each qubit to contribute to multiple basis functions exponentially. This inversion resolves the contradiction by achieving higher representational capacity without increasing the quantity of qubits.
3Measurement precision
If the quantum circuit uses more encoding repetitions, then training accuracy may improve, but the training time and computational resources increase
Solution Approach 1:
The patent changes the parameter of rotation angles by introducing scaling factors, which allows the quantum circuit to achieve better training accuracy more efficiently. The scaled angles enable the variational quantum gates to operate in a more expressive parameter space, improving the quality of learned representations without requiring proportionally more training iterations or time, thus resolving the contradiction between measurement precision and loss of time.
Data Source
AI summary
A system and method for encoding a dataset in a quantum circuit for quantum machine learning in a system includes providing a dataset comprising a plurality of input features; for each input feature of the plurality of input features, applying the plurality of encoding quantum gates on one quantum bit (qubit) or a plurality of qubits, wherein each of the plurality of encoding quantum gates rotates the one qubit or the plurality of qubits by a rotation angle which is proportional to the input feature and one of a plurality of scaling factors, each of the plurality of encoding quantum gates is assigned a different one of the plurality of scaling factors, and the plurality of scaling factors comprises powers of two; applying the plurality of variational quantum gates; determining a plurality of measurement values for the qubit; adjusting the quantum circuit; and determining output data.