Asset allocation method and device, electronic equipment, storage medium and program product

Abstract

The invention provides an asset allocation method and device, and the method comprises the steps: receiving an asset allocation task, carrying out the quantum coding of the asset allocation task, and forming quantum coding information used for representing the task, and the quantum coding information comprises quantum bit information and quantum energy information; determining an initial quantum state of the quantum bit based on the quantum bit information; a noisy quantum circuit is adopted, the initial quantum state is controlled and processed based on the quantum energy information, a noisy measurement result is obtained, and the noisy quantum circuit comprises a quantum circuit which uses noise to execute a quantum optimization algorithm; performing de-noising processing on the noisy measurement result to obtain a noise-free measurement result; and determining an asset configuration result based on the noise-free measurement result. According to the invention, the noise influence of quantum computing equipment can be effectively relieved, the performance of a quantum algorithm is improved, and the accuracy and robustness of asset allocation are improved.

Description

Technical Field

[0001] The present disclosure relates to the field of quantum information, specifically to technical fields such as quantum computing and smart finance, and in particular to an asset allocation method, device, electronic device, storage medium, and program product. Background Art

[0002] Asset allocation is a core problem in modern finance. Finding the optimal balance between expected return and risk to achieve the optimal asset portfolio is crucial. Traditional asset allocation algorithms rely on classical computational techniques such as linear programming, quadratic programming, and Monte Carlo simulation. However, as the number of asset portfolio options increases, these problems quickly become extremely complex and computationally intensive, especially when considering multiple constraints and complex financial models.

[0003] In recent years, quantum computing has been proposed as a potential tool to address these types of problems, potentially more efficient than classical computing for certain types of computation. In particular, variational quantum algorithms such as the Quantum Approximate Optimization Algorithm (QAOA) and the Quantum Variational Eigensolver (VQE) have been used to solve portfolio optimization problems in finance.

[0004] Although quantum approximate optimization algorithms and quantum variational eigensolvers have been used for financial optimization problems, these methods do not directly deal with the noise problem of quantum hardware, resulting in high error rates, insufficient accuracy and reliability in quantum algorithms. Summary of the Invention

[0005] The present disclosure provides an asset allocation method, apparatus, electronic device, storage medium, and program product to improve the accuracy and robustness of asset allocation.

[0006] According to one aspect of the present disclosure, an asset allocation method is provided, comprising: receiving an asset allocation task, quantum encoding the asset allocation task to form quantum coded information for representing the task, wherein the quantum coded information includes quantum bit information and quantum energy information; determining an initial quantum state of the quantum bit based on the quantum bit information; using a noisy quantum circuit to control the initial quantum state based on the quantum energy information to obtain a noisy measurement result, wherein the noisy quantum circuit includes a quantum circuit that uses noise to execute a quantum optimization algorithm; denoising the noisy measurement result to obtain a noise-free measurement result; and determining an asset allocation result based on the noise-free measurement result.

[0007] Optionally, the asset allocation task includes: the number of assets to be configured, an optimization objective, and constraints; the quantum encoding of the asset allocation task to form quantum coded information for representing the task includes: obtaining the quantum bit information based on the number of assets to be configured; obtaining a problem Hamiltonian based on the optimization objective; obtaining a constrained Hamiltonian based on the constraints; obtaining a cost Hamiltonian based on the problem Hamiltonian and the constrained Hamiltonian; constructing a mixed Hamiltonian; and using the cost Hamiltonian and the mixed Hamiltonian as the quantum energy information.

[0008] Optionally, the method further includes: creating a noise model based on a preset error probability; and applying the noise model to the quantum circuit that executes the quantum optimization algorithm to obtain the noisy quantum circuit.

[0009] Optionally, the noise model includes a depolarization noise model configured to introduce noise with the error probability; applying the noise model to the quantum circuit that executes the quantum optimization algorithm to obtain the noisy quantum circuit includes: applying the noise introduced by the depolarization noise model to all quantum gates of the quantum circuit that executes the quantum optimization algorithm to obtain the noisy quantum circuit.

[0010] Optionally, the use of a noisy quantum circuit to control and process the initial quantum state based on the quantum energy information to obtain a noisy measurement result includes: using a noisy quantum circuit to evolve and measure the initial quantum state based on the quantum energy information; if an end condition is met, using the measurement result when the end condition is met as the noisy measurement result; wherein, meeting the end condition includes: the measurement result converges, or reaching a preset number of iterations.

[0011] Optionally, the noisy quantum circuit includes: a parameterized layer having adjustable parameters; the evolving and measuring the initial quantum state based on the quantum energy information using the noisy quantum circuit includes: evolving a current input state based on the adjustable parameters and the quantum energy information using the parameterized layer to obtain a current output state; the initial value of the current input state is the initial quantum state; measuring the current output state to obtain a current measurement result; if the end condition is not met, updating the adjustable parameters based on the current measurement result, and re-inputting the current output state into the noisy quantum circuit until the end condition is met.

[0012] Optionally, the noise includes multiple levels of noise, and the noisy measurement results include multiple groups of noisy measurement results, each group of noisy measurement results corresponding to one level of noise; the denoising processing of the noisy measurement results to obtain noise-free measurement results includes: performing zero-noise extrapolation processing on the multiple groups of noisy measurement results to obtain the noise-free measurement results.

[0013] Optionally, performing zero-noise extrapolation processing on the multiple groups of noisy measurement results includes: performing zero-noise extrapolation processing on the multiple groups of noisy measurement results using at least one of linear extrapolation, polynomial extrapolation and exponential extrapolation.

[0014] According to another aspect of the present disclosure, an asset allocation device is provided, comprising: a conversion module for receiving an asset allocation task, performing quantum encoding on the asset allocation task, and forming quantum encoding information for representing the task; the quantum encoding information comprising quantum bit information and quantum energy information; a generation module for determining an initial quantum state of the quantum bit based on the quantum bit information; a processing module for controlling the initial quantum state using a noisy quantum circuit based on the quantum energy information to obtain a noisy measurement result; wherein the noisy quantum circuit comprises a quantum circuit that uses noise to execute a quantum optimization algorithm; a denoising module for denoising the noisy measurement result to obtain a noise-free measurement result; and a determination module for determining an asset allocation result based on the noise-free measurement result.

[0015] According to another aspect of the present disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute any one of the methods described in any one of the above aspects.

[0016] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute any one of the methods according to any one of the above aspects.

[0017] According to another aspect of the present disclosure, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements the method according to any one of the above aspects.

[0018] In an embodiment of the present disclosure, the initial quantum state is controlled and processed by a noisy quantum circuit, the noisy measurement result is denoised, and a noise-free measurement result is obtained. The asset allocation result is determined based on the noise-free measurement result. The present disclosure can determine a noise-free result that is close to the ideal situation from the quantum computing result affected by noise. This can effectively alleviate the impact of noise in quantum computing devices (such as medium-scale quantum devices), improve the performance of quantum algorithms, and thus improve the accuracy and robustness of asset allocation.

[0019] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings are used to better understand the present invention and do not constitute a limitation of the present invention.

[0021] Figure 1 is a schematic diagram according to a first embodiment of the present disclosure;

[0022] Figure 2 is a schematic diagram of a noisy quantum approximate optimization algorithm circuit provided according to an embodiment of the present disclosure;

[0023] Figure 3 is a schematic diagram of performing zero-noise extrapolation processing on multiple sets of noisy measurement results according to some embodiments of the present disclosure;

[0024] Figure 4 is a schematic diagram according to a second embodiment of the present disclosure;

[0025] Figure 5 Schematic diagram of an electronic device used to implement the asset configuration method according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0026] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0027] In the existing technology, although quantum approximate optimization algorithms and quantum variational eigenvalue solvers have been used for financial optimization problems, since quantum approximate optimization algorithms and quantum variational eigenvalue solvers do not directly deal with the noise problem of quantum hardware, the effect of using quantum approximate optimization algorithms or quantum variational eigenvalue solvers for financial asset portfolio optimization needs to be improved. Specifically, the accuracy, stability and application scope all need to be improved.

[0028] To improve the performance of quantum algorithms and the accuracy of asset allocation, this paper combines a denoising algorithm (using zero-noise extrapolation as an example) with a quantum optimization algorithm, specifically applying zero-noise extrapolation to asset portfolio optimization. By introducing noise into the quantum computation process and then applying zero-noise extrapolation to determine the result under noise-free conditions, this approach can significantly improve the accuracy and robustness of quantum algorithms.

[0029] Figure 1 This is a schematic diagram of the first embodiment of the present disclosure. This embodiment provides an asset allocation method, such as Figure 1 As shown, the method includes:

[0030] Step 101: Receive an asset allocation task, perform quantum encoding on the asset allocation task, and form quantum encoding information for representing the task, wherein the quantum encoding information includes quantum bit information and quantum energy information.

[0031] Step 102: Determine the initial quantum state of the quantum bit based on the quantum bit information.

[0032] Step 103: Using a noisy quantum circuit to control and process the initial quantum state based on the quantum energy information to obtain a noisy measurement result, wherein the noisy quantum circuit includes a quantum circuit that uses noise to execute a quantum optimization algorithm.

[0033] Step 104: De-noise the noisy measurement result to obtain a noise-free measurement result.

[0034] Step 105: Determine an asset allocation result based on the noise-free measurement result.

[0035] The present disclosure uses a noisy quantum circuit to control the initial quantum state, denoises the noisy measurement results, obtains noise-free measurement results, and determines asset allocation results based on the noise-free measurement results. The present disclosure can determine a noise-free result that is close to the ideal situation from quantum computing results affected by noise. This can effectively alleviate the impact of noise in quantum computing devices (such as medium-scale quantum devices), improve the performance of quantum algorithms, and thus improve the accuracy and robustness of asset allocation.

[0036] In some embodiments of the present disclosure, the asset allocation task includes: the number of assets to be allocated, an optimization objective, and constraints. The quantum encoding of the asset allocation task to form quantum coded information for representing the task includes:

[0037] Obtaining the quantum bit information based on the number of assets to be configured;

[0038] Obtaining the problem Hamiltonian based on the optimization objective;

[0039] Obtaining a constrained Hamiltonian based on the constraint conditions;

[0040] Obtaining a cost Hamiltonian based on the problem Hamiltonian and the constraint Hamiltonian;

[0041] Constructing mixed Hamiltonians;

[0042] The cost Hamiltonian and the mixed Hamiltonian are used as the quantum energy information.

[0043] In some embodiments of the present disclosure, in order to employ quantum computing to solve asset allocation tasks, it is necessary to convert the classical optimization problem into a quantum computing problem. Step 101 further includes: qubit encoding and objective function encoding, and, if constraints exist, constraint encoding. Qubit encoding obtains qubit information; objective function encoding converts the classical objective function into a target quantity processed by quantum computing; and constraint encoding converts the classical constraints into constraints processed by quantum computing, with the target and constraints being quantum energy information.

[0044] In qubit encoding, a qubit is the basic unit of information in quantum computing. Quantum gates can operate on one, two, or more qubits. The quantum state of a qubit is used to represent the weight of each asset. This allows asset allocation results to be derived based on the quantum state of the qubit.

[0045] When the number of assets to be configured is n, then at least set qubits to represent all asset portfolios. Indicates rounding up. For example, if the total number of assets is 4, then at least 2 qubits should be set.

[0046] The quantum state of each quantum bit is used to characterize whether the corresponding asset is configured, such as using |1> to indicate configuration and |0> to indicate non-configuration.

[0047] For example, assuming that the number of assets to be configured is n=4, then 4 quantum bits are prepared, and the quantum states of each quantum bit are |1>, |0>, |1>, |0>, respectively, which means: the first and third assets are configured, and the second and fourth assets are not configured.

[0048] The allocation ratio of each asset can be determined based on a set rule. For example, if the rule is equal weight allocation, the weights of the asset portfolio after allocation are (0.5, 0, 0.5, 0). The above is just an example. In actual resource allocation scenarios, the allocation ratio of different assets is often influenced by a variety of factors such as expected returns, risk, personal preferences, and market environment. Therefore, specific rules can be set to determine the allocation in actual situations.

[0049] For the objective function and constraint encoding, in the asset allocation scenario, the optimization objectives include maximizing the expected return under a given risk level, or minimizing the risk under a given expected return.

[0050] The expected return can be expressed as:

[0051]

[0052] Where R is the expected return; w i is the weight of the i-th asset; r i is the rate of return of the i-th asset; n is the number of assets to be allocated.

[0053] Risk is measured by the variance between the returns of assets and can be expressed as:

[0054]

[0055] Among them, σ 2 is a risk metric; σ ij is the covariance between the return rate of the ith asset and the return rate of the jth asset; w i is the weight of the i-th asset; w j is the weight of the j-th asset; n is the number of assets to be configured.

[0056] Furthermore, in asset allocation scenarios, there are usually some constraints, such as the sum of asset weights being equal to 1.

[0057] In order to use quantum computing to solve the optimization problem, the above-mentioned classical objective function and constraints are converted into the amount of computation processed by quantum computing to obtain quantum energy information, which is, for example, the Hamiltonian.

[0058] Taking the quantum circuit as an example, the Hamiltonian includes: cost Hamiltonian H C and the mixed Hamiltonian H M .

[0059] Cost Hamiltonian H CIt is determined according to the optimization goal and constraints. Specifically, the problem Hamiltonian H can be determined according to the above optimization goal, and the constraint Hamiltonian H can be determined according to the constraints. constraint , and then based on the problem Hamiltonian H and the constraint Hamiltonian H constraint Determine the cost Hamiltonian H C .

[0060] The Hamiltonian H can be expressed as:

[0061] H=H R -λH σ

[0062] Among them, H R is the term related to expected return; H σ It is a risk-related item; specifically, it is a set adjustment coefficient, which can be any integer, usually a value between 0 and 1.

[0063] H R It can be expressed as:

[0064]

[0065] Among them, r i is the rate of return of the ith asset; Z i is the Pauli Z operator acting on the quantum bit corresponding to the i-th asset; n is the number of assets to be configured.

[0066] H σ It can be expressed as:

[0067]

[0068] Among them, σ ij is the covariance between the return rate of the ith asset and the return rate of the jth asset; Z i is the Pauli Z operator acting on the quantum bit corresponding to the i-th asset; Z j is the Pauli Z operator acting on the quantum bit corresponding to the j-th asset; n is the number of assets to be configured.

[0069] Constrained Hamiltonian H constraint It can be expressed as:

[0070]

[0071] Among them, Z iis the Pauli Z operator applied to the qubit corresponding to the i-th asset; n is the number of assets to be allocated; and N is the desired sum of weights, such as N = 1 or N = n. N = 1 indicates that the quantum state of the entire system (the state of each qubit or the sum of its combinations) reflects a sum of weights equal to 1, meaning that all weights add up to 100%. N = n indicates that all assets are fully invested, with each asset's weight being maximized. For example, the sum of all asset weights equals 1, and each asset is equally weighted. Thus, the energy of this term is minimized only when the sum of all weights satisfies the constraints.

[0072] Based on the above problem Hamiltonian H and constraint Hamiltonian H constraint , the cost Hamiltonian H can be determined C , expressed as:

[0073] H C =-H+H constraint =-(H R -λH σ )+H constraint

[0074] Mixed Hamiltonian H M is an inherent part of the quantum approximate optimization algorithm, used to search the solution space, and can be expressed as:

[0075]

[0076] Among them, X i is the Pauli X operator acting on the quantum bit corresponding to the i-th asset.

[0077] Through quantum coding, the asset allocation task is converted into quantum coding information, which includes quantum bit information (quantum state) and quantum energy information. Quantum energy information is, for example, Hamiltonian. Then, quantum calculation is performed based on the quantum bit information and Hamiltonian to obtain the asset allocation result.

[0078] In some embodiments of the present disclosure, in step 102, quantum coding may include quantum bit coding, objective function coding, and constraint condition coding. Quantum bit information is obtained through quantum bit coding, and quantum energy information, such as Hamiltonian, is obtained through objective function coding and constraint condition coding.

[0079] The qubit information may include the number n of qubits. Based on this number n, the initial quantum state of these n qubits can be generated. For example, this initial quantum state can be obtained through quantum state preparation. Quantum state preparation includes initializing the quantum state of the qubits for subsequent quantum operations and calculations.

[0080] In quantum computing, quantum state preparation initializes the quantum state of each qubit to a specific initial quantum state. In quantum algorithms such as quantum approximate optimization, the initial quantum state places all qubits in a superposition state to search all possible solutions. During the initial quantum state preparation phase, although each qubit is physically only one qubit, it can simultaneously represent multiple states through quantum state superposition. By initializing each qubit to a superposition state, the entire quantum computing device can search all possible solutions in parallel, facilitating quantum computing acceleration.

[0081] Specifically, in a quantum simulator (using Qiskit as an example), initializing the simulation involves setting simulator parameters, such as the number of qubits. Code is then written to simulate quantum gate operations, such as Hadamard gates (H-gates) and other quantum gates, to generate the initial quantum state. In other words, a quantum simulator can be used to simulate quantum gate operations based on the number of qubits to obtain the initial quantum state of the qubits.

[0082] In some embodiments of the present disclosure, in order to solve the problem of inaccurate asset allocation results caused by the quantum approximate optimization algorithm and the quantum variational eigensolver not directly processing the noise of the quantum hardware, in step 103, additional noise is introduced into the quantum circuit. The quantum circuit that uses the noise to execute the quantum optimization algorithm is a noisy quantum circuit, and the measurement result obtained based on the noisy quantum circuit is a noisy measurement result.

[0083] The method may further include: creating a noise model based on a preset error probability; and applying the noise model to the quantum circuit executing the quantum optimization algorithm to obtain the noisy quantum circuit. After creating the noise model and its corresponding error probability (e.g., 0.01), noise is automatically introduced during quantum gate operations in the quantum circuit, such as by changing execution time or interactions between qubits.

[0084] Furthermore, by setting multiple error probabilities, such as 0.01, 0.05, etc., different levels of noise can be introduced, and multiple groups of noisy measurement results can be obtained in subsequent processing.

[0085] Furthermore, the noise model may be a depolarization noise model, which is configured to introduce noise with a preset error probability. Based on this, the noise introduced by the depolarization noise model is applied to all quantum gates of the quantum circuit executing the quantum optimization algorithm to obtain the noisy quantum circuit. The depolarization noise randomly changes the state of the quantum bit to an arbitrary state, causing the quantum bit to uniformly become a mixture of all possible states, thereby reducing the purity of the information.

[0086] In some embodiments of the present disclosure, in step 103, the use of a noisy quantum circuit to control the initial quantum state based on the quantum energy information to obtain a noisy measurement result includes:

[0087] Using a noisy quantum circuit, evolving and measuring the initial quantum state based on the quantum energy information;

[0088] If the end condition is met, the measurement result when the end condition is met is used as the noisy measurement result;

[0089] The reaching of the termination condition includes: the measurement result converges, or the reaching of a preset number of iterations.

[0090] After obtaining the noisy quantum circuit, the initial quantum state of the quantum bit is used as the input state of the noisy quantum circuit. The measurement result of the quantum bit is obtained through the evolution and measurement of the noisy quantum circuit.

[0091] The above evolution and measurement may be repeated multiple times until the above end condition is met, and the measurement result when the end condition is met is used as the noisy measurement result.

[0092] In some embodiments of the present disclosure, the noisy quantum circuit includes: a parameterized layer having adjustable parameters;

[0093] The adopting of a noisy quantum circuit to evolve and measure the initial quantum state based on the quantum energy information includes:

[0094] Using the parameterized layer, evolving a current input state based on the adjustable parameter and the quantum energy information to obtain a current output state; the initial value of the current input state is the initial quantum state;

[0095] Measuring the current output state to obtain a current measurement result;

[0096] If the termination condition is not met, the adjustable parameter is updated based on the current measurement result, and the current output state is re-input into the noisy quantum circuit until the termination condition is met.

[0097] Taking the quantum approximate optimization algorithm as an example, the noisy quantum circuit is a noisy quantum approximate optimization algorithm circuit. Figure 2 Schematic diagram of a noisy quantum approximate optimization algorithm circuit provided according to an embodiment of the present disclosure.

[0098] like Figure 2 As shown, the noisy quantum approximate optimization algorithm circuit includes: an initialization layer 201, a parameterization layer 202 and a measurement layer 203.

[0099] Initialization layer 201 is used to obtain the initial quantum state of the qubits. Referring to the description of step 102 above, a superposition state can be generated using a Hadamard gate (H gate) based on quantum encoding information (e.g., the number of qubits) as the initial quantum state. To this end, the initialization layer can obtain the initial quantum state generated by the above process.

[0100] The parameterized layer 202 has adjustable parameters and is used to evolve the initial quantum state of the qubit based on the adjustable parameters and quantum energy information to obtain the final state of the qubit. The parameterized layer consists of a series of parameterized quantum gates whose parameters are updateable. By optimizing these parameters, the final quantum state can provide an approximate solution to the optimization problem. The parameterized layer is typically multi-layered, with the number of layers (or depth) represented by p. p can be set based on computational accuracy and complexity requirements.

[0101] In the quantum approximate optimization algorithm, the quantum energy information includes the cost Hamiltonian and the mixed Hamiltonian. Each parameterized layer includes the cost operator and the mixed operator, which are expressed as:

[0102]

[0103]

[0104] Among them, U(H C ,γ) is the cost operator; U(H M ,β) is the mixing operator; H C is the cost Hamiltonian; H M is a mixed Hamiltonian; β, γ are a set of updateable parameters corresponding to each layer (depth), i.e., adjustable parameters.

[0105] During execution, the cost operator and the hybrid operator are used alternately to evolve the quantum state of the qubit. The evolution and measurement can be performed multiple times, and each time the parameterized layer evolves the current input state to obtain the current output state. Specifically, the parameterized layer evolves the quantum state based on adjustable parameters (β, γ) and quantum energy information (the cost Hamiltonian and the hybrid Hamiltonian).

[0106] The adjustable parameters of the parameterized layer are updatable. Specifically, they can be updated based on measurement results and classical optimization algorithms (such as gradient descent and simulated annealing) to minimize the expected cost of the quantum circuit measurement results. This evolution and parameter update process is repeated until a termination condition is reached, such as convergence to a satisfactory solution or reaching a preset number of iterations. The solution corresponding to the measurement result when the termination condition is reached is regarded as the optimal solution or a near-optimal solution.

[0107] To improve accuracy and robustness, some embodiments of the present disclosure introduce noise into quantum circuits, specifically into the parameterized layer. For example, noise can be introduced by varying the execution time of quantum gates or by changing the interactions between quantum bits.

[0108] The parameterized layer before the introduction of noise is represented by U, and the parameterized layer after the introduction of noise is represented by U * Said that, for this reason, Figure 2 As shown, the p groups of parameterized layers are expressed as:

[0109]

[0110] The measurement layer 203 measures the final state of the qubits and obtains the measurement result. After measurement, the quantum state of each qubit collapses to a specific classical bit, such as 0 or 1. After evolution and measurement in the quantum circuit, a classical bit string, such as "0101," is obtained.

[0111] In some embodiments of the present disclosure, in step 104, the noise-containing measurement result is subjected to denoising to obtain a noise-free measurement result. The noise includes multiple levels of noise, and the noisy measurement result includes multiple groups of noisy measurement results, each group of noisy measurement results corresponding to a level of noise.

[0112] This step outputs the measurement results in the presence of noise. By comparing the measurement results with and without noise, or with different noise levels, we can determine how noise affects the output of the quantum computation. Specifically, we can simulate different levels of noise by changing the value of the error probability (e.g., 0.02, 0.03, etc.) and repeating the above process. By collecting and analyzing the results under different noise levels, we can determine the impact of noise on the computation results.

[0113] De-noising is performed on the noisy measurement results to obtain noise-free measurement results, including: performing zero-noise extrapolation processing on multiple groups of noisy measurement results to obtain noise-free measurement results.

[0114] Zero Noise Extrapolation (ZNE) performs calculations at varying noise levels and extrapolates the data to determine the results under noise-free conditions. Using ZNE, noisy measurements can be denoised to produce noise-free results.

[0115] Perform zero-noise extrapolation on multiple sets of noisy measurement results, including:

[0116] At least one of linear extrapolation, polynomial extrapolation and exponential extrapolation is used to perform zero-noise extrapolation on multiple groups of noisy measurement results.

[0117] refer to Figure 3 , Figure 3 Schematic diagram of performing zero-noise extrapolation on multiple sets of noisy measurement results according to some embodiments of the present disclosure.

[0118] This step exploits the relationship between noise level and computational error to recover the ideal noise-free result from noisy data. The following are the detailed steps for using zero-noise extrapolation:

[0119] Step 301: Collect data at different noise levels.

[0120] Specifically, the noise intensity is systematically increased in the quantum simulator, and output data is collected under each noise level to obtain the above-mentioned multiple sets of noisy measurement results.

[0121] Step 302: Select an extrapolation method.

[0122] Zero-noise extrapolation is used to extrapolate noisy data to zero noise. Extrapolation methods include linear, polynomial, and exponential. The choice depends on the noise characteristics and experimental data. Linear extrapolation assumes the results vary linearly with the noise intensity; polynomial extrapolation is suitable when the results have a nonlinear relationship with the noise intensity; and exponential extrapolation is used when the data exhibit exponential dependence.

[0123] Step 303: Extrapolate to zero noise.

[0124] Based on the selected extrapolation method, the collected data is used to determine the results under noise-free conditions. Specifically, by fitting the above data collected under different noise levels, including:

[0125] Each noise level and the corresponding measurement result are considered as data points;

[0126] Fit these data using linear regression, polynomial fitting, or other statistical methods;

[0127] Based on the fitted curve, determine the result when the noise level is zero.

[0128] Step 304: Result analysis and verification.

[0129] The extrapolated zero-noise estimate is used to improve the quantum computation result. The validity and accuracy of the extrapolated result can then be verified by comparing it with known results or benchmark results.

[0130] As the noise level increases, the measurement results decrease linearly. Linear regression can be used to determine the results at zero noise. By employing zero-noise extrapolation, we can determine a near-ideal result from quantum computations affected by noise, which helps improve the performance of quantum algorithms in quantum devices.

[0131] In some embodiments, a noise-free measurement result can be obtained through zero-noise extrapolation, and an asset allocation result can be determined based on the noise-free measurement result. The noise-free measurement result is decoded to obtain the asset allocation result. Specifically, the asset allocation result can be represented by asset portfolio weights. For example, if the total number of assets is 4, the noise-free measurement result is 1010, and the preset allocation rule is equal weight allocation, then after decoding, the asset allocation result is represented by the asset portfolio weights as (0.5, 0, 0.5, 0).

[0132] In addition, the decoding results can be analyzed and verified, and the results that meet the requirements are used as the final asset allocation results. Specifically, the results obtained by zero-noise extrapolation are analyzed and compared with those obtained by traditional computer financial models or classical optimization algorithms to evaluate the accuracy and robustness of the quantum algorithm and ensure that the investment portfolio provided by the quantum algorithm is the optimal choice. The following are the detailed steps for result analysis and verification:

[0133] The first step is to aggregate and process the results.

[0134] Collect and organize all calculation results, including those without noise, with noise, and after applying the zero-noise extrapolation method. This includes data cleaning, formatting, and preliminary statistical analysis. Data cleaning ensures that all results are recorded according to the same standards and format; preliminary statistical analysis calculates basic indicators such as the mean and standard deviation of the results in various cases.

[0135] The second step is to compare and evaluate the results.

[0136] Results corrected using the zero-noise extrapolation method should be compared with other data points (e.g., ideal results in the absence of noise, and noisy but uncorrected results). This evaluates whether the zero-noise extrapolation method effectively reduces the impact of noise and improves the accuracy of the results. Results comparison compares the corrected results with the ideal results or baseline values, while effectiveness evaluation analyzes the impact of the zero-noise extrapolation method on improving the accuracy of the results and assesses its effectiveness.

[0137] The third step is to conduct statistical analysis.

[0138] Conduct in-depth statistical analysis on different data sets (e.g., original data and corrected data at different noise levels), including hypothesis testing, confidence interval calculation, and correlation analysis. Hypothesis testing is used to test whether there is a significant difference between the corrected results and the ideal results; confidence intervals are calculated for the corrected results to assess the uncertainty and credibility of the results.

[0139] The fourth step is verification and cross-validation.

[0140] Use independent datasets or methods to verify the accuracy of the zero-noise extrapolation correction results. This is achieved by comparing the results on different algorithms or different hardware. Specifically, independent verification is performed by repeating the calculation on different quantum hardware or using different quantum algorithms to verify the consistency and reliability of the results. Cross-validation, for example, by using other error mitigation techniques or theoretical models as comparisons, further verifies the effectiveness of the zero-noise extrapolation method.

[0141] These steps can ensure the accuracy and reliability of quantum computing results and fully evaluate the effectiveness of error mitigation strategies, such as zero-noise extrapolation, which will help enhance the practical feasibility and trustworthiness of quantum computing applications.

[0142] The present disclosure obtains noise-free measurement results by denoising noisy measurement results, and obtains asset allocation results based on the noise-free measurement results. It can determine noise-free results close to the ideal situation from quantum computing results affected by noise, which can effectively alleviate the impact of noise in quantum computing devices (such as medium-scale quantum devices), improve the performance of quantum algorithms, and thus improve the accuracy and robustness of asset allocation.

[0143] Furthermore, in noisy quantum computing environments, traditional quantum algorithms may be unable to consistently find optimal solutions due to the influence of noise, especially when dealing with complex optimization problems such as portfolio optimization. This invention, by introducing noise and denoising the noisy measurement results, not only improves the performance of quantum algorithms but also enhances their robustness in real-world noisy environments. Even in noisy conditions, quantum algorithms are able to search the valid solution space and, through denoising, approach the optimal solution in a noise-free environment, thus improving the stability of quantum computing.

[0144] While conventional quantum algorithms are limited in their practicality in financial applications due to noise and computational errors, the quantum algorithms disclosed in this paper can provide more reliable and accurate results, thereby expanding the application of quantum computing in financial optimization and other complex computational problems.

[0145] Therefore, the resource allocation method of this embodiment has improvements in accuracy, stability and application scope, thereby improving the asset allocation effect.

[0146] Figure 4 4 is a schematic diagram of a second embodiment of the present disclosure, which provides an asset allocation device. The asset allocation device 400 includes a conversion module 401, a generation module 402, a processing module 403, a denoising module 404, and a determination module 405.

[0147] The conversion module 401 is configured to receive an asset allocation task, perform quantum encoding on the asset allocation task, and form quantum coded information configured to represent the task, where the quantum coded information includes quantum bit information and quantum energy information;

[0148] The generation module 402 is configured to determine an initial quantum state of the qubit based on the qubit information;

[0149] The processing module 403 is configured to use a noisy quantum circuit to control and process the initial quantum state based on quantum energy information to obtain a noisy measurement result, wherein the noisy quantum circuit includes a quantum circuit configured to execute a quantum optimization algorithm;

[0150] The denoising module 404 is configured to perform denoising on the noisy measurement results to obtain noise-free measurement results;

[0151] The determination module 405 is configured to determine an asset allocation result based on the noise-free measurement result.

[0152] In some embodiments, the asset allocation task includes: the number of assets to be allocated, the optimization goal, and the constraints;

[0153] The conversion module 401 is further configured to:

[0154] Based on the number of assets to be configured, obtain quantum bit information;

[0155] Obtain the Hamiltonian of the problem based on the optimization goal;

[0156] Obtain the constrained Hamiltonian based on the constraint conditions;

[0157] Based on the problem Hamiltonian and the constraint Hamiltonian, the cost Hamiltonian is obtained;

[0158] Constructing mixed Hamiltonians;

[0159] The cost Hamiltonian and the mixed Hamiltonian are used as quantum energy information.

[0160] In some embodiments, the apparatus 400 further includes:

[0161] Add a module configured to create a noise model based on a preset error probability; configure the noise model to execute the quantum circuit of the quantum optimization algorithm to obtain a noisy quantum circuit.

[0162] In some embodiments, the noise model includes a depolarization noise model configured to introduce noise with an error probability;

[0163] Add module further configuration as:

[0164] The noise introduced by the depolarization noise model should be configured in all quantum gates of a quantum circuit that executes a quantum optimization algorithm to obtain a noisy quantum circuit.

[0165] In some embodiments, the processing module 403 is further configured to:

[0166] Using noisy quantum circuits, the initial quantum state is evolved and measured based on quantum energy information;

[0167] If the end condition is met, the measurement result at the time the end condition is met will be regarded as the noisy measurement result;

[0168] The end condition is reached when the measurement result converges or when a preset number of iterations is reached.

[0169] In some embodiments, a noisy quantum circuit includes: a parameterized layer having adjustable parameters;

[0170] The processing module 403 is further configured to:

[0171] Using a parameterized layer, the current input state is evolved based on adjustable parameters and quantum energy information to obtain the current output state; the initial value of the current input state is the initial quantum state;

[0172] Measure the current output state and obtain the current measurement result;

[0173] If the termination condition is not met, the adjustable parameters are updated based on the current measurement results, and the current output state is re-input into the noisy quantum circuit until the termination condition is met.

[0174] In some embodiments, the noise includes multiple levels of noise, and the noisy measurement results include multiple groups of noisy measurement results, each group of noisy measurement results corresponding to a level of noise;

[0175] The denoising module 404 is further configured to:

[0176] De-noising is performed on noisy measurement results to obtain noise-free measurement results, including:

[0177] The zero-noise extrapolation method is used to process multiple groups of noisy measurement results to obtain noise-free measurement results.

[0178] In some embodiments, the denoising module 404 is further configured to:

[0179] At least one of linear extrapolation, polynomial extrapolation and exponential extrapolation is used to perform zero-noise extrapolation on multiple groups of noisy measurement results.

[0180] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0181] Figure 5 A schematic block diagram of an example electronic device 500 that can be used to implement embodiments of the present disclosure is shown. The electronic device 500 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are provided as examples only and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0182] like Figure 5 As shown, the electronic device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. Various programs and data required for the operation of the electronic device 500 can also be stored in the RAM 503. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0183] Multiple components in the electronic device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a magnetic disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the electronic device 500 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0184] The computing unit 501 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 501 performs the various methods and processes described above, such as the asset allocation method. For example, in some embodiments, the asset allocation method can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 508. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 500 via the ROM 502 and / or the communication unit 509. When the computer program is loaded into the RAM 503 and executed by the computing unit 501, one or more steps of the asset allocation method described above can be performed. Alternatively, in other embodiments, the computing unit 501 can be configured to perform the asset allocation method by any other suitable means (e.g., via firmware).

[0185] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0186] The program code for implementing the method of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0187] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0188] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0189] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0190] A computer system may include a client and a server. The client and server are generally remote from each other and typically interact via a communication network. This client-server relationship is established by computer programs running on the respective computers, establishing a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host, a host product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosts and VPS services ("Virtual Private Servers" or simply "VPS"). The server may also be a server in a distributed system or a server integrated with blockchain.

[0191] It should be understood that the various forms of the processes shown above can be used to reorder, add, or remove steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved. This is not a limitation herein.

[0192] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

[0193] It can be understood that in the embodiments of the present disclosure, the same or similar contents in different embodiments can be referenced to each other.

[0194] It can be understood that the terms “first”, “second”, etc. in the embodiments of the present disclosure are only used for distinction and do not indicate the degree of importance, time sequence, etc.

[0195] It is understandable that, unless otherwise specified, the order of the steps in the process indicates that the temporal relationship between these steps is not limited.

[0196] In the technical solutions disclosed herein, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

Claims

1. An asset allocation method, characterized in that: include: receiving an asset allocation task, performing quantum encoding on the asset allocation task to form quantum coded information representing the task, wherein the quantum coded information includes quantum bit information and quantum energy information; determining an initial quantum state of the quantum bit based on the quantum bit information; Using a noisy quantum circuit to control and process the initial quantum state based on the quantum energy information to obtain a noisy measurement result, wherein the noisy quantum circuit includes a quantum circuit that uses noise to execute a quantum optimization algorithm; Performing denoising on the noisy measurement result to obtain a noise-free measurement result; An asset allocation result is determined based on the noise-free measurement result.

2. The method according to claim 1, characterized in that The asset allocation task includes: the number of assets to be allocated, the optimization target and the constraints; The quantum encoding of the asset allocation task to form quantum coded information for representing the task includes: Obtaining the quantum bit information based on the number of assets to be configured; Obtaining the problem Hamiltonian based on the optimization objective; Obtaining a constrained Hamiltonian based on the constraint conditions; Obtaining a cost Hamiltonian based on the problem Hamiltonian and the constraint Hamiltonian; Constructing mixed Hamiltonians; The cost Hamiltonian and the mixed Hamiltonian are used as the quantum energy information.

3. The method according to claim 1, characterized in that Also includes: Create a noise model based on the preset error probability; The noise model is applied to the quantum circuit that executes the quantum optimization algorithm to obtain the noisy quantum circuit.

4. The method according to claim 3, characterized in that The noise model includes a depolarization noise model configured to introduce noise with the error probability; The step of applying the noise model to the quantum circuit that executes the quantum optimization algorithm to obtain the noisy quantum circuit comprises: The noise introduced by the depolarization noise model is applied to all quantum gates of the quantum circuit that executes the quantum optimization algorithm to obtain the noisy quantum circuit.

5. The method according to claim 1, wherein The method of using a noisy quantum circuit to control and process the initial quantum state based on the quantum energy information to obtain a noisy measurement result includes: Using a noisy quantum circuit, evolving and measuring the initial quantum state based on the quantum energy information; If the end condition is met, the measurement result when the end condition is met is used as the noisy measurement result; The reaching of the termination condition includes: the measurement result converges, or the reaching of a preset number of iterations.

6. The method according to claim 5, characterized in that The noisy quantum circuit includes: a parameterized layer having adjustable parameters; The adopting of a noisy quantum circuit to evolve and measure the initial quantum state based on the quantum energy information includes: Using the parameterized layer, evolving a current input state based on the adjustable parameter and the quantum energy information to obtain a current output state; the initial value of the current input state is the initial quantum state; Measuring the current output state to obtain a current measurement result; If the termination condition is not met, the adjustable parameter is updated based on the current measurement result, and the current output state is re-input into the noisy quantum circuit until the termination condition is met.

7. The method according to claim 1, characterized in that The noise includes multiple levels of noise, and the noisy measurement results include multiple groups of noisy measurement results, each group of noisy measurement results corresponding to one level of noise; The denoising process is performed on the noisy measurement result to obtain a noise-free measurement result, including: The multiple groups of noisy measurement results are processed using a zero-noise extrapolation method to obtain the noise-free measurement results.

8. The method according to claim 7, characterized in that The performing zero-noise extrapolation processing on the multiple groups of noisy measurement results includes: At least one of linear extrapolation, polynomial extrapolation and exponential extrapolation is used to perform zero-noise extrapolation on the multiple groups of noisy measurement results.

9. An asset allocation device, characterized in that: include: a conversion module configured to receive an asset allocation task, perform quantum encoding on the asset allocation task, and form quantum coded information configured to represent the task, wherein the quantum coded information includes quantum bit information and quantum energy information; A generating module configured to determine an initial quantum state of the quantum bit based on the quantum bit information; a processing module configured to use a noisy quantum circuit to control and process the initial quantum state based on the quantum energy information to obtain a noisy measurement result; wherein the noisy quantum circuit includes a quantum circuit that uses noise to execute a quantum optimization algorithm; a denoising module configured to perform denoising on the noisy measurement result to obtain a noise-free measurement result; A determination module is configured to determine an asset allocation result based on the noise-free measurement result.

10. An electronic device comprising: at least one processor; as well as A memory communicatively connected to the at least one processor; characterized in that The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 8.

11. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-8.

12. A computer program product, characterized in that The invention comprises a computer program which, when executed by a processor, implements the method according to any one of claims 1 to 8.

Images