Node combination method, apparatus, device, medium, and product

Abstract

This disclosure relates to the field of computer technology, and in particular to a method, apparatus, device, medium, and product for node combination. In this disclosure, each node in a set to be processed is quantum-state encoded to obtain its quantum ground state. Based on the quantum ground state of each node, with the objective of minimizing the grouping error loss, the logical state of each node is determined. The logical state characterizes the grouping probability of a node relative to a candidate subset, and the grouping error loss characterizes the grouping error probability of two connected nodes. Based on the logical state of each node, a target combination is determined in the set to be processed. By using a small amount of qubit resources to encode the quantum state of each node, the logical state of each node that can characterize the grouping probability is determined in iterative optimization. The target combination can then be determined through the logical state, turning the determination of node combination into a continuous optimization problem, thus achieving node combination optimization. Therefore, the accuracy of node combination optimization can be improved.

Description

Technical Field

[0001] This disclosure belongs to the field of computer technology, and specifically relates to a node combination method, apparatus, device, medium and product. Background Technology

[0002] Typically, the partitioning of node combinations can be viewed as a combinatorial optimization problem, a typical Boolean satisfiability problem. As the problem size increases, the number of possible solutions grows exponentially, making the search for the optimal solution extremely difficult. In the quantum realm, QAOA (Quantum Approximate Optimization Algorithm) and quantum annealing are the most commonly used methods for solving combinatorial optimization problems, demonstrating good performance in terms of solution quality.

[0003] In related technologies, QAOA requires the use of a general-purpose quantum computer. Due to hardware bottlenecks, the demand for quantum resources (such as the number of qubits and quantum gates) increases dramatically with the complexity of the problem, leading to error accumulation, increased control difficulty, and ever-increasing requirements for error calibration and cooling, sometimes even making it impossible to complete the most basic solutions. Quantum annealing algorithms mostly employ optical quantum coherent Ising machines, whose optical parametric oscillators have a rapidly increasing complexity with problem size, potentially increasing the difficulty of physical implementation. Therefore, this reduces the accuracy of node combinatorial optimization. Summary of the Invention

[0004] This disclosure addresses some deficiencies mentioned in the background art by providing a node combination method, apparatus, device, medium, and product that can improve the accuracy of node combination optimization.

[0005] In a first aspect, embodiments of this disclosure provide a node combination method, comprising: Quantum state encoding is performed on each node in the set to be processed to obtain the quantum ground state of each node; Based on the quantum ground state of each node, with the goal of minimizing the grouping error loss, the logical state of each node is determined. The logical state is used to characterize the grouping probability of the node relative to the candidate subset, and the grouping error loss is used to characterize the grouping error probability of two nodes that are connected. Based on the logical state of each node, a target combination is determined in the set to be processed.

[0006] Optionally, the step of quantum state encoding for each node in the set to be processed to obtain the quantum ground state of each node includes: Based on the number of nodes, multiple qubits are determined, and the number of nodes is logarithmically related to the number of qubits. Quantum state encoding is performed on the multiple qubits to obtain the quantum ground state of each node.

[0007] Optionally, determining the logical state of each node based on its quantum ground state, with the goal of minimizing grouping error loss, includes: Based on the quantum ground state of each node, determine the logic state to be determined for each node; Based on the undetermined logical state of any two nodes and the grouping error loss function, the grouping error loss of the two nodes is determined, and the grouping error loss function is used to maximize the connection relationship between each node; With the goal of minimizing the grouping error loss, the parameterized quantum circuit is optimized until the output of the parameterized quantum circuit is a preset target variable quantum state, thereby obtaining the logic state of each node.

[0008] Optionally, determining the undetermined logical state of each node based on its quantum ground state includes: The parameterized quantum circuit is used to quantum encode each of the quantum ground states to obtain the current circuit variational parameters of each node; The undetermined logic state of each node is determined based on the square of the absolute value of the variational parameter of the current circuit.

[0009] Optionally, the parameterized quantum circuit includes at least one of the following layers: The input layer is used to input the quantum ground states. Quantum gates are used to quantum encode the quantum ground state to obtain the variational parameters of the current circuit; The output layer is used to determine and output the variational quantum state based on the variational parameters of the current circuit.

[0010] Optionally, the grouping error loss function represents the sum of the squared terms of the first sub-function and the squared terms of the second sub-function; The first sub-function represents the difference between the absolute value of the sum of the logical states of any two nodes and the preset grouping threshold. The second sub-function represents the difference between the absolute value of the difference between the logical states of any two nodes and the preset grouping threshold.

[0011] Optionally, the target combination includes a first candidate subset and a second candidate subset, and determining the target combination from the set to be processed based on the logical state of each node includes: Based on the logical states and preset grouping thresholds, the nodes in the set to be processed are divided. When the logical state is greater than the preset grouping threshold, it is determined that the node belongs to the first candidate subset in the target combination; When the logical state is less than or equal to the preset grouping threshold, the node is determined to belong to the second candidate subset in the target combination.

[0012] Optionally, the node is a user in the social network; the first candidate subset is the set of nodes selected as seed nodes in the social network, where the seed node represents the starting point of information propagation; and the second candidate subset is the set of other nodes in the social network.

[0013] Optionally, the preset grouping threshold is negatively correlated with the number of nodes.

[0014] In a second aspect, embodiments of this disclosure provide a node assembly apparatus, comprising: The encoding module is used to encode the quantum state of each node in the set to be processed, so as to obtain the quantum ground state of each node; The first determining module is used to determine the logical state of each node based on the quantum ground state of each node, with the goal of minimizing the grouping error loss. The logical state is used to characterize the grouping probability of the node relative to the candidate subset, and the grouping error loss is used to characterize the grouping error probability of two nodes that are connected. The second determining module is used to determine the target combination in the set to be processed based on the logical state of each node.

[0015] In a third aspect, embodiments of this disclosure provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described node combination method.

[0016] In a fourth aspect, embodiments of this disclosure provide a computer-readable storage medium having a computer program stored thereon, the program being executed by a processor to implement the above-described node combination method.

[0017] In a fifth aspect, embodiments of this disclosure provide a computer program product including computer-readable code, or a non-volatile computer-readable storage medium carrying computer-readable code, wherein when the computer-readable code is run in a processor of an electronic device, the processor in the electronic device executes the above-described node combination method.

[0018] In this disclosure, each node in the set to be processed is quantum-state encoded to obtain its quantum ground state. Based on the quantum ground state of each node, the logical state of each node is determined with the objective of minimizing the grouping error loss. The logical state characterizes the grouping probability of a node relative to a candidate subset, and the grouping error loss characterizes the grouping error probability of two connected nodes. Based on the logical state of each node, a target combination is determined in the set to be processed. By using a small amount of qubit resources to encode the quantum state of each node, the logical state of each node that can characterize the grouping probability is determined in iterative optimization, and the target combination can be determined through the logical state, thus achieving combinatorial optimization of nodes. In this way, without embedding graphical instances into the quantum circuit, the discrete Boolean satisfiability problem of combinatorial optimization is transformed into a continuous optimization problem, and it can be extended to other combinatorial optimization problems with similar structures, reducing the error rate of optimization solutions. Therefore, the accuracy of node combinatorial optimization can be improved.

[0019] It should be understood that both the foregoing general description and the following detailed description are exemplary and intended to provide further illustration of the claimed technology. Attached Figure Description

[0020] Figure 1 This is a flowchart of a node combination method provided in this disclosure.

[0021] Figure 2 A schematic diagram of the parameterized quantum circuit provided in this disclosure.

[0022] Figure 3 Another flowchart is provided for a node combination method disclosed herein.

[0023] Figure 4 This is a schematic diagram of the original topology connections of each node in the set to be processed provided in this disclosure.

[0024] Figure 5 This is a histogram of the logical state of each node in the combination to be processed provided in this disclosure.

[0025] Figure 6 This is a schematic diagram of the node combination after the nodes in the set to be processed are divided, as provided in this disclosure.

[0026] Figure 7 This is a schematic diagram of the structure of a node combination device provided in this disclosure.

[0027] Figure 8 This is a hardware block diagram of an electronic device provided in this disclosure.

[0028] Figure 9 This is a schematic diagram of a computer program product provided in this disclosure. Detailed Implementation

[0029] To enable those skilled in the art to better understand the technical solution of this application, the application scenario of this application will be described first below.

[0030] Typically, the partitioning of node combinations can be viewed as a combinatorial optimization problem, a typical Boolean satisfiability problem. As the problem size increases, the number of possible solutions grows exponentially, making the search for the optimal solution extremely difficult. In the quantum realm, QAOA (Quantum Approximate Optimization Algorithm), VQE (Variational Quantum Eigensolver), and quantum annealing are the most commonly used methods for solving combinatorial optimization problems, demonstrating good performance in terms of solution quality. QAOA and VQE both involve designing hybrid quantum-classical algorithms to gradually approach the global optimum through iterative optimization of parameterized quantum circuits; while quantum annealing utilizes the principle of adiabatic evolution to find the ground state corresponding to the objective Hamiltonian, which is the optimal solution.

[0031] In related technologies, QAOA requires the use of a general-purpose quantum computer. Due to hardware bottlenecks, the demand for quantum resources (such as the number of qubits and quantum gates) increases dramatically with the complexity of the problem, leading to error accumulation, increased control difficulty, and ever-increasing requirements for error calibration and cooling, sometimes even making it impossible to complete the most basic solutions. Quantum annealing algorithms mostly employ optical quantum coherent Ising machines, whose optical parametric oscillators have a rapidly increasing complexity with problem size, potentially increasing the difficulty of physical implementation. Therefore, this reduces the accuracy of node combinatorial optimization.

[0032] Furthermore, all of the above schemes use base encoding, that is, the computation of the base state for each qubit ( or Encoding the logical state of the corresponding problem variables means that the number of qubits required is linearly related to the size of the problem. Specifically, the combinatorial optimization problem discussed in this paper requires N qubits to solve for the optimal value under this encoding, while current quantum computers can only support a few hundred bits and face severe fault tolerance challenges. This limits the use of the above algorithm to very small problem instances under current conditions.

[0033] To address the aforementioned technical problems, this disclosure provides an inventive concept: quantum state encoding is performed on each node using a small amount of qubit resources. During iterative optimization, the logical state of each node, which can characterize the grouping probability, is determined. Furthermore, the target combination can be determined through the logical state, achieving combinatorial optimization of the nodes. In this way, without embedding graphical instances into the quantum circuit, the discrete Boolean satisfiability problem of combinatorial optimization is transformed into a continuous optimization problem, and it can be extended to other combinatorial optimization problems with similar structures, reducing the error rate of the optimization solution. Therefore, the accuracy of node combinatorial optimization can be improved.

[0034] The present disclosure will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present disclosure and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the present disclosure are shown in the drawings, not the entire structure.

[0035] Figure 1 This is a flowchart illustrating a node combination method provided in this disclosure. Figure 1 As shown, the method includes: S101: Encode the quantum state of each node in the set to be processed to obtain the quantum ground state of each node.

[0036] Specifically, in this embodiment, the nodes in the set to be processed are quantized and mapped to the corresponding quantum ground state. This allows for quantum computing using parameterized quantum circuits, enabling the search for the global optimal solution through combinatorial optimization methods, thereby determining the node combination.

[0037] S102: Based on the quantum ground state of each node, determine the logical state of each node with the goal of minimizing the grouping error loss.

[0038] Specifically, multiple qubits are used to quantum encode the quantum ground state. This embodiment defines variable parameters in the parameterized quantum circuit by defining variable quantum states. With the goal of minimizing grouping error loss, the parameterized quantum circuit is run multiple times to determine the logic state of each node. This eliminates the need to encode graphical instances of the connection relationships between nodes in the set to be processed into the parameterized quantum circuit; it only requires minimizing the grouping error loss to solve the combinatorial optimization problem. Here, the logic state characterizes the grouping probability of a node relative to a candidate subset, and the grouping error loss characterizes the grouping error probability of two connected nodes. This encoding method encodes the logic state of each node into the same Hilbert space, mapping each node's logic state to a range of probability values ​​rather than a specific probability value, thus fully utilizing Hilbert's solution space.

[0039] S103: Based on the logical state of each node, determine the target combination in the set to be processed.

[0040] Specifically, the logical state of each node is judged to divide the nodes in the set to be processed, and the final division result is taken as the optimal solution of the combinatorial optimization problem, i.e., the target combinatorial.

[0041] In this disclosure, each node in the set to be processed is quantum-state encoded to obtain its quantum ground state. Based on the quantum ground state of each node, the logical state of each node is determined with the objective of minimizing the grouping error loss. The logical state characterizes the grouping probability of a node relative to a candidate subset, and the grouping error loss characterizes the grouping error probability of two connected nodes. Based on the logical state of each node, a target combination is determined in the set to be processed. By using a small amount of qubit resources to encode the quantum state of each node, the logical state of each node that can characterize the grouping probability is determined in iterative optimization, and the target combination can be determined through the logical state, thus achieving combinatorial optimization of nodes. In this way, without embedding graphical instances into the quantum circuit, the discrete Boolean satisfiability problem of combinatorial optimization is transformed into a continuous optimization problem, and it can be extended to other combinatorial optimization problems with similar structures, reducing the error rate of optimization solutions. Therefore, the accuracy of node combinatorial optimization can be improved.

[0042] In one possible implementation, an exemplary method for quantum-encoding each node in the set to be processed to obtain the quantum ground state of each node includes: Based on the number of nodes, multiple qubits are determined; quantum state encoding is performed on multiple qubits to obtain the quantum ground state of each node.

[0043] Specifically, in this embodiment, the set to be processed has N nodes, and the number of nodes is logarithmically related to the number of qubits. That is, only using By reducing the number of qubits exponentially, the number of quantum ground states at each node is reduced. These quantum ground states serve as inputs to parameterized quantum circuits, allowing the solution to be accomplished using shallower quantum circuits.

[0044] For example, in a set of nodes to be processed, each node i is mapped to its corresponding quantum ground state. .for For a given number of qubits, a variable quantum state is defined as a variable parameter of the parameterized quantum circuit to determine the logic state of each node. An exemplary expression for the variable quantum state is as follows:

[0045] in, For variational parameters, For the number of qubits, For rotation angle, It is the quantum ground state. For variational parameters.

[0046] In one possible implementation, an exemplary method for determining the logical state of each node based on its quantum ground state, with the goal of minimizing grouping error loss, includes: Based on the quantum ground state of each node, the undetermined logical state of each node is determined; based on the undetermined logical states of any two nodes and the grouping error loss function, the grouping error loss of the two nodes is determined; with the goal of minimizing the grouping error loss, the parameterized quantum circuit is optimized until the output of the parameterized quantum circuit is the preset target variable quantum state, thus obtaining the logical state of each node.

[0047] Specifically, in this embodiment, node combination refers to dividing multiple nodes in the set to be processed. These nodes may be connected, and the goal is to divide all nodes into two subsets, maximizing the sum of the number of edges between nodes in the different subsets. This is similar to the maximum cut problem, thus achieving the target node combination partition. Therefore, the grouping error loss function is used to maximize the connections between nodes.

[0048] In this embodiment, we do not need to encode the graphical instances of the node connections into the gate circuits of the parameterized quantum circuit. Therefore, we only need to construct a reasonable loss function based on the problem to solve the combinatorial optimization problem. Throughout the optimization process, the parameterized quantum circuit is executed and measured multiple times to obtain a reliable logic state.

[0049] For example, based on the quantum ground state of each node, the undetermined logical state of each node is determined, including: The parameterized quantum circuit is used to quantum encode each quantum ground state to obtain the current circuit variational parameters of each node; based on the square of the absolute value of the current circuit variational parameters, the undetermined logic state of each node is determined.

[0050] Specifically, each quantum ground state is input into a parameterized quantum circuit, where it is encoded through multiple quantum gates in a quantum gate layer to obtain the variational parameters of the current circuit. The square of the variational parameters of the current circuit is calculated to obtain the logic state to be determined. An exemplary formula for determining the logic state is as follows:

[0051] in, For rotation angle, This is a logical state.

[0052] During the optimization process, the variational parameters of the circuit can be updated using a classic optimizer. The grouping error loss function is optimized until convergence, that is, the grouping error loss is minimized until the output variational parameters are defined as variational quantum states, thereby determining the logical state of each node.

[0053] For example, a parameterized quantum circuit includes at least one of the following layers: The input layer is used to input each quantum ground state; the quantum gate layer is used to quantum encode the quantum ground state to obtain the variational parameters of the current circuit; the output layer is used to determine and output the variational quantum state based on the variational parameters of the current circuit.

[0054] Figure 2 A schematic diagram of the parameterized quantum circuit provided in this disclosure, such as Figure 2 As shown, each quantum ground state is input into a parameterized quantum circuit, where it is encoded through multiple quantum gates in a quantum gate layer to output the variational parameters of the current circuit. The parameterized quantum circuit is executed multiple times to make the variational parameters of the current circuit the defined target variational quantum state. Finally, the variational quantum state is output through an output layer.

[0055] This parameterized quantum circuit has sufficient characterization power to approach the optimal solution in Hilbert space. Figure 2 The figure represents a parameterized quantum circuit with one layer when n is 2, where each layer consists of 2N single-qubit gates and N CNOT gates. Furthermore, better experimental results can be obtained by increasing the number of layers in the circuit.

[0056] Quantum gate multiplexing techniques typically perform well with simple circuits by pre-measuring and reusing idle circuits. However, gates for solving production problems are often intricately entangled. Re-encoding and constructing Hamiltonians requires building unique quantum circuits for each specific problem model, and this process is mathematically complex, leading to increasing circuit depth. In this embodiment, reducing the number of qubits by an order of magnitude effectively addresses the issue of circuit depth.

[0057] For example, the grouping error loss function represents the sum of the squared terms of the first sub-function and the squared terms of the second sub-function; the first sub-function represents the difference between the absolute value of the sum of the logical states of any two nodes and the preset grouping threshold; the second sub-function represents the difference between the absolute value of the difference between the logical states of any two nodes and the preset grouping threshold.

[0058] Specifically, by minimizing the loss function, the optimal quantum state of the solution to this problem can be found. The construction of the grouping error loss function is explained below. First, determine all pairs of nodes j and k in the set to be processed that are connected by an edge; define two sub-functions. and ,in , These represent the logical states of nodes j and k, respectively. Using the two sub-functions mentioned above, the first sub-function can be constructed. Second sub-function Thus, the grouping error loss function is obtained. An exemplary formula for the grouping error loss function is as follows:

[0059] in, As the first sub-function, For the second sub-function, The preset grouping threshold is defined, and j and k are two nodes connected by an edge.

[0060] By minimizing the grouping error loss function, the two squared terms will approach 0, at which point the expression will be satisfied. and Both are true, so we can obtain and 0 and or And 0. Based on the preset grouping threshold, the probability distributions of nodes j and k are on both sides of the threshold. Therefore, there exist node pairs j and k connected by edges that are partitioned into the left and right subsets. Considering all node pairs, the optimal solution can be obtained, yielding the target node combination.

[0061] In one possible implementation, the target combination includes a first candidate subset and a second candidate subset. An exemplary method for determining the target combination from the set to be processed based on the logical state of each node includes: Based on each logical state and a preset grouping threshold, each node in the set to be processed is divided; when the logical state is greater than the preset grouping threshold, the node is determined to belong to the first candidate subset of the target combination; when the logical state is less than or equal to the preset grouping threshold, the node is determined to belong to the second candidate subset of the target combination.

[0062] Specifically, the logical state is a probability, with high and low probabilities representing the first and second candidate subsets, respectively. In one possible implementation, the preset grouping threshold is negatively correlated with the number of nodes.

[0063] The following explains the preset grouping threshold. In this embodiment, assuming that low-probability positions represent left nodes and high-probability positions represent right nodes, an intermediate probability threshold can be set so that the logical state corresponds to a range of probabilities rather than a specific probability value. This allows full utilization of Hilbert's solution space. An exemplary formula for presetting the grouping threshold is as follows:

[0064] in, To preset the grouping threshold, This indicates the number of nodes in the first candidate subset on the left.

[0065] Thus, when the logical state is greater than If node i is not in the first candidate subset on the left, then node i is assigned to the second candidate subset on the right; otherwise, node i is assigned to the second candidate subset on the right. However, It is unknown before the experiment begins, therefore it can be iterated over. All possible values ​​were finally determined. The initial value is defined as That is, the preset grouping threshold is set to .

[0066] In one possible implementation, the node is a user in the social network; the first candidate subset is the set of nodes in the social network selected as seed nodes, which represent the starting point of information propagation; and the second candidate subset is the set of other nodes in the social network.

[0067] Specifically, this embodiment uses maximizing the influence of social networks as a specific application scenario. The problem is characterized by how to select a small subset of nodes (called seed nodes) within a social network as the starting point for information dissemination, so that the information can reach as many network users as possible. The basic idea is to divide the users in the network into two subsets: the first candidate subset represents users selected as seed nodes, and the second candidate subset consists of other node users who were not selected. The goal is to maximize the number of connections between these two subsets of users through social relationships, corresponding to maximizing the potential for information dissemination.

[0068] Figure 3 Another flowchart is provided for a node combination method according to this disclosure. Figure 3 As shown, the method includes: S301: Define the variable quantum state.

[0069] Specifically, the set to be processed has N nodes, and each node i is mapped to the corresponding quantum ground state. .for For a given qubit, a variable quantum state is defined. Exemplary expressions for the variable quantum state are as shown above and will not be repeated here.

[0070] S302: Set the preset grouping threshold and grouping error loss function.

[0071] Specifically, iterate through all possible values ​​of the node to be assigned to the first candidate subset on the left, and finally determine the preset grouping threshold as the reciprocal of the number of nodes to be processed. The grouping error loss function represents the sum of the squared terms of the first sub-function and the squared terms of the second sub-function; the first sub-function represents the difference between the absolute value of the sum of the logical states of any two nodes and the preset grouping threshold; the second sub-function represents the difference between the absolute value of the difference of the logical states of any two nodes and the preset grouping threshold.

[0072] S303: Run the parameterized quantum circuit multiple times to obtain the logic state of the node.

[0073] Specifically, each quantum ground state is input into a parameterized quantum circuit, where it is encoded through multiple quantum gates in a quantum gate layer to output the variational parameters of the current circuit. The parameterized quantum circuit is executed multiple times to make the variational parameters of the current circuit the defined target variational quantum state. Finally, the variational quantum state is output through the output layer, thereby calculating and determining the logic state of each node.

[0074] S304: Divide the target node combination according to the preset grouping threshold.

[0075] Specifically, the logical state of each node is judged by using a preset grouping threshold. Nodes with logical states greater than the preset grouping threshold are divided into the first candidate subset, and those with logical states less than or equal to the preset grouping threshold are divided into the second candidate subset, thereby determining the target node combination.

[0076] For example, taking node 4 as an example, using two qubits to encode the quantum state, the variable quantum state is obtained as follows:

[0077] Set the preset grouping threshold to 1 / 4, divide the logical state of each node, and determine the encoding. and Two nodes whose logical state probabilities are greater than the preset grouping threshold are assigned to the first candidate subset on the left, while the rest are assigned to the second candidate subset on the right.

[0078] Figure 4 This is a schematic diagram of the original topological connections of each node in the set to be processed provided in this disclosure, such as... Figure 4 As shown. Encoded as The nodes and their encodings are The nodes and encoding are The nodes have a connection relationship, encoded as The nodes and encoding are The nodes and encoding are The nodes have a connection relationship, encoded as The nodes and encoding are The nodes and encoding are The nodes have a connection relationship, encoded as The nodes and encoding are The nodes and encoding are The nodes have a connection relationship.

[0079] Figure 5 The logical state histogram of each node in the combination to be processed provided in this disclosure is as follows: Figure 5 As shown. Encoded as The logical state of a node is greater than a preset grouping threshold, and is encoded as follows. The logical state of the node is less than the preset grouping threshold, and it is encoded as follows. The logical state of a node is greater than a preset grouping threshold, and is encoded as follows. The logical state of the node is less than the preset grouping threshold.

[0080] Figure 6 This is a schematic diagram of the node combinations after partitioning the nodes in the set to be processed, as provided in this disclosure. Figure 6 As shown. The final target node combination can be obtained as follows, encoded as... The nodes and encoding are The node belongs to the first candidate subset and is encoded as The nodes and encoding are The node belongs to the second candidate subset.

[0081] In one possible application scenario, users are designated as nodes, and social relationships between users are designated as edges. All nodes are then divided into two candidate subsets, maximizing the sum of the number of edges between nodes in different subsets, similar to the maximum cut problem. Traditional quantum methods construct a QUBO (Quadratic Unconstrained Binary Optimization) model for this problem, then transform it into the Ising Hamiltonian of the objective function based on variable mapping, and finally design corresponding parameterized gate circuits for iterative optimization. This disclosure employs a unique encoding method that does not require embedding graphical instances into the quantum circuit. It only requires designing a loss function based on the problem's objective value for iterative parameter optimization, thus ensuring the stability of the proposed quantum gate circuit. By encoding all states into the same Hilbert space, the logical state of each node is mapped to a probability range, transforming the discrete Boolean satisfiability problem into a continuous optimization problem, achieving... This quantum circuit solves N-node problems using a single qubit and can be extended to combinatorial optimization problems with similar structures, enabling the solution of problems involving thousands of qubits—a feat previously impossible on comparable quantum computers. Furthermore, the shallower depth of the quantum circuit reduces noise sensitivity and the error rate during the solution process. The grouping error loss function is designed as a continuous mean square error estimator, moving away from the traditional VQA technique of Hamiltonian expectation, and ultimately measuring the corresponding computational state.

[0082] Furthermore, combinatorial optimization problems are ubiquitous in daily life, closely related to logistics, investment, drug development, materials, scheduling, and fault detection. Classical computers often struggle to solve large-scale problems of this kind due to insufficient computing power. Quantum computers, with their superposition and quantum entanglement properties, possess powerful parallel processing capabilities. However, in the early stages of quantum development, hardware limitations hindered the performance of quantum algorithms. On the one hand, the expensive qubit resources required for large-scale problems increase the cost of research and development; on the other hand, using complex quantum circuits in such cases often introduces stronger noise and higher error rates. Therefore, the qubit reduction technology proposed in this proposal significantly saves quantum resources and greatly expands the application scenarios of quantum algorithms.

[0083] Figure 7 This is a schematic diagram of a node assembly device provided in this disclosure. Figure 7 As shown, the device 700 includes: an encoding module 710, a first determination module 720, and a second determination module 730.

[0084] The encoding module 710 is used to encode the quantum state of each node in the set to be processed, so as to obtain the quantum ground state of each node; The first determining module 720 is used to determine the logical state of each node based on the quantum ground state of each node, with the goal of minimizing the grouping error loss. The logical state is used to characterize the grouping probability of the node relative to the candidate subset, and the grouping error loss is used to characterize the grouping error probability of two nodes that are connected. The second determining module 730 is used to determine the target combination in the set to be processed based on the logical state of each node.

[0085] Optionally, the encoding module is used for: Based on the number of nodes, multiple qubits are determined, and the number of nodes is logarithmically related to the number of qubits. Quantum state encoding is performed on the multiple qubits to obtain the quantum ground state of each node.

[0086] Optionally, the first determining module is used to: Based on the quantum ground state of each node, determine the logic state to be determined for each node; Based on the undetermined logical state of any two nodes and the grouping error loss function, the grouping error loss of the two nodes is determined, and the grouping error loss function is used to maximize the connection relationship between each node; With the goal of minimizing the grouping error loss, the parameterized quantum circuit is optimized until the output of the parameterized quantum circuit is a preset target variable quantum state, thereby obtaining the logic state of each node.

[0087] Optionally, the first determining module is used to: The parameterized quantum circuit is used to quantum encode each of the quantum ground states to obtain the current circuit variational parameters of each node; The undetermined logic state of each node is determined based on the square of the absolute value of the variational parameter of the current circuit.

[0088] Optionally, the parameterized quantum circuit includes at least one of the following layers: The input layer is used to input the quantum ground states. Quantum gates are used to quantum encode the quantum ground state to obtain the variational parameters of the current circuit; The output layer is used to determine and output the variational quantum state based on the variational parameters of the current circuit.

[0089] Optionally, the grouping error loss function represents the sum of the squared terms of the first sub-function and the squared terms of the second sub-function; The first sub-function represents the difference between the absolute value of the sum of the logical states of any two nodes and the preset grouping threshold. The second sub-function represents the difference between the absolute value of the difference between the logical states of any two nodes and the preset grouping threshold.

[0090] Optionally, the target combination includes a first candidate subset and a second candidate subset, and the second determining module is used to: Based on the logical states and preset grouping thresholds, the nodes in the set to be processed are divided. When the logical state is greater than the preset grouping threshold, it is determined that the node belongs to the first candidate subset in the target combination; When the logical state is less than or equal to the preset grouping threshold, the node is determined to belong to the second candidate subset in the target combination.

[0091] Optionally, the node is a user in the social network; the first candidate subset is the set of nodes selected as seed nodes in the social network, where the seed node represents the starting point of information propagation; and the second candidate subset is the set of other nodes in the social network.

[0092] Optionally, the preset grouping threshold is negatively correlated with the number of nodes.

[0093] This application also provides an electronic device for performing the above-described node combination method. Please refer to... Figure 8 It illustrates a schematic diagram of an electronic device provided by some embodiments of this application. For example... Figure 8 As shown, the electronic device 80 includes: a processor 800, a memory 801, a bus 802, and a communication interface 803. The processor 800, the communication interface 803, and the memory 801 are connected via the bus 802. The memory 801 stores a computer program that can run on the processor 800. When the processor 800 runs the computer program, it executes the node combination method provided in any of the foregoing embodiments of this application.

[0094] The memory 801 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this device network element and at least one other network element is achieved through at least one communication interface 803 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc.

[0095] Bus 802 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. The memory 801 is used to store programs. After receiving an execution instruction, the processor 800 executes the program. The node combination method disclosed in any of the foregoing embodiments of this application can be applied to the processor 800, or implemented by the processor 800.

[0096] The processor 800 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 800 or by instructions in software form. The processor 800 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules may reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 801. Processor 800 reads the information in memory 801 and, in conjunction with its hardware, completes the steps of the above method.

[0097] The electronic device provided in this application embodiment and the node combination method provided in this application embodiment are based on the same inventive concept and have the same beneficial effects as the methods they adopt, operate or implement.

[0098] This application also provides a computer-readable storage medium corresponding to the node combination method provided in the foregoing embodiments. The computer-readable storage medium shown may be an optical disc, on which a computer program is stored. When the computer program is run by a processor, it executes the node combination method provided in any of the foregoing embodiments.

[0099] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical and magnetic storage media, which will not be elaborated here.

[0100] The computer-readable storage medium provided in the above embodiments of this application and the node combination method provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the applications stored therein.

[0101] This application also provides a computer program product 900, such as... Figure 9 As shown. This computer program product carries a computer program 901. The instructions included in the program code can be used to execute the steps of the node combination method described in the above method embodiments. For details, please refer to the above method embodiments, which will not be repeated here.

[0102] The aforementioned computer program product can be implemented through hardware, software, or a combination thereof. In one optional embodiment, the computer program product is specifically embodied in a computer storage medium; in another optional embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.

[0103] The basic principles of this disclosure have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this disclosure are merely examples and not limitations, and should not be considered as essential features of each embodiment of this disclosure. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the scope of this disclosure to the necessity of employing the aforementioned specific details for implementation.

[0104] The block diagrams of devices, apparatuses, devices, and systems disclosed herein are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0105] Additionally, as used herein, the "or" used in a list of items beginning with "at least one" indicates a separate list, such that a list of, for example, "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the word "exemplary" does not imply that the described example is preferred or better than other examples.

[0106] It should also be noted that in the systems and methods of this disclosure, the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered as equivalent solutions to this disclosure.

[0107] Various changes, substitutions, and modifications can be made to the technology described herein without departing from the teachings defined by the appended claims. Furthermore, the scope of the claims of this disclosure is not limited to the specific aspects of the processes, machines, manufactures, events, means, methods, and actions described above. Currently existing or later-developed processes, machines, manufactures, events, means, methods, or actions that perform substantially the same function or achieve substantially the same result as the corresponding aspects described herein can be utilized. Therefore, the appended claims include such processes, machines, manufactures, events, means, methods, or actions within their scope.

[0108] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of this disclosure. Therefore, this disclosure is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features disclosed herein.

[0109] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this disclosure to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A node combination method, characterized in that, include: Quantum state encoding is performed on each node in the set to be processed to obtain the quantum ground state of each node; Based on the quantum ground state of each node, with the goal of minimizing the grouping error loss, the logical state of each node is determined. The logical state is used to characterize the grouping probability of the node relative to the candidate subset, and the grouping error loss is used to characterize the grouping error probability of two nodes that are connected. Based on the logical state of each node, a target combination is determined in the set to be processed.

2. The method according to claim 1, characterized in that, The process of quantum-encoding each node in the set to be processed to obtain the quantum ground state of each node includes: Based on the number of nodes, multiple qubits are determined, and the number of nodes is logarithmically related to the number of qubits. Quantum state encoding is performed on the multiple qubits to obtain the quantum ground state of each node.

3. The method according to claim 1, characterized in that, The determination of the logical state of each node based on its quantum ground state, with the goal of minimizing grouping error loss, includes: Based on the quantum ground state of each node, determine the logic state to be determined for each node; Based on the undetermined logical state of any two nodes and the grouping error loss function, the grouping error loss of the two nodes is determined, and the grouping error loss function is used to maximize the connection relationship between each node; With the goal of minimizing the grouping error loss, the parameterized quantum circuit is optimized until the output of the parameterized quantum circuit is a preset target variable quantum state, thereby obtaining the logic state of each node.

4. The method according to claim 3, characterized in that, The process of determining the undetermined logical state of each node based on its quantum ground state includes: The parameterized quantum circuit is used to quantum encode each of the quantum ground states to obtain the current circuit variational parameters of each node; The undetermined logic state of each node is determined based on the square of the absolute value of the variational parameter of the current circuit.

5. The method according to any one of claims 3 or 4, characterized in that, The parameterized quantum circuit includes at least one of the following layers: The input layer is used to input the quantum ground states. Quantum gates are used to quantum encode the quantum ground state to obtain the variational parameters of the current circuit; The output layer is used to determine and output the variational quantum state based on the variational parameters of the current circuit.

6. The method according to claim 3, characterized in that, The grouping error loss function represents the sum of the squared terms of the first sub-function and the squared terms of the second sub-function; The first sub-function represents the difference between the absolute value of the sum of the logical states of any two nodes and the preset grouping threshold. The second sub-function represents the difference between the absolute value of the difference between the logical states of any two nodes and the preset grouping threshold.

7. The method according to claim 1, characterized in that, The target combination includes a first candidate subset and a second candidate subset. Determining the target combination from the set to be processed based on the logical state of each node includes: Based on the logical states and preset grouping thresholds, the nodes in the set to be processed are divided. When the logical state is greater than the preset grouping threshold, it is determined that the node belongs to the first candidate subset in the target combination; When the logical state is less than or equal to the preset grouping threshold, the node is determined to belong to the second candidate subset in the target combination.

8. The method according to claim 7, characterized in that, The node is a user in the social network; the first candidate subset is the set of nodes selected as seed nodes in the social network, where the seed node represents the starting point of information propagation; and the second candidate subset is the set of other nodes in the social network.

9. The method according to claim 7, characterized in that, The preset grouping threshold is negatively correlated with the number of nodes.

10. A node assembly device, characterized in that, include: The encoding module is used to encode the quantum state of each node in the set to be processed, so as to obtain the quantum ground state of each node; The first determining module is used to determine the logical state of each node based on the quantum ground state of each node, with the goal of minimizing the grouping error loss. The logical state is used to characterize the grouping probability of the node relative to the candidate subset, and the grouping error loss is used to characterize the grouping error probability of two nodes that are connected. The second determining module is used to determine the target combination in the set to be processed based on the logical state of each node.

Images