CN117591947B - Data classification method of quantum support vector machine based on variable component sub-core - Google Patents
Data classification method of quantum support vector machine based on variable component sub-core Download PDFInfo
- Publication number
- CN117591947B CN117591947B CN202410071683.3A CN202410071683A CN117591947B CN 117591947 B CN117591947 B CN 117591947B CN 202410071683 A CN202410071683 A CN 202410071683A CN 117591947 B CN117591947 B CN 117591947B
- Authority
- CN
- China
- Prior art keywords
- data
- mapping
- quantum
- representing
- classified
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000012706 support-vector machine Methods 0.000 title claims abstract description 38
- 238000013507 mapping Methods 0.000 claims abstract description 148
- 230000006870 function Effects 0.000 claims abstract description 71
- 238000000926 separation method Methods 0.000 claims abstract description 25
- 239000011159 matrix material Substances 0.000 claims description 21
- 238000002360 preparation method Methods 0.000 claims description 8
- 238000012360 testing method Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000013506 data mapping Methods 0.000 claims description 2
- 230000007423 decrease Effects 0.000 claims description 2
- 230000008569 process Effects 0.000 abstract description 5
- 238000009826 distribution Methods 0.000 abstract description 3
- 230000000694 effects Effects 0.000 abstract description 3
- 238000010801 machine learning Methods 0.000 abstract description 2
- 238000005457 optimization Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000005305 interferometry Methods 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Medical Informatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to the field of quantum machine learning, in particular to a data classification method of a quantum support vector machine based on a variable component sub-core. The method solves the problem that the data classification method based on the quantum support vector machine is inaccurate in classification effect when complex data classification is processed. The method specifically comprises the following steps: mapping the data set to a mapping space through a mapping function, and then carrying out quantum coding; solving a quantum state representing the mapping space data separation hyperplane parameters; classifying the mapping data set by using a quantum support vector machine; updating the mapping function based on the loss function to obtain an optimized mapping function; and mapping the data to be classified based on the optimized mapping function, and classifying the data by using a quantum support vector machine. The invention maps the data to be classified into the high-dimensional mapping space and uses hyperplane separation, thereby enabling the quantum support vector machine to process more complex data distribution and improving the accuracy of classification.
Description
Technical Field
The invention relates to the field of quantum machine learning, in particular to a data classification method of a quantum support vector machine based on a variable component sub-core.
Background
Complex data refers to data that cannot be separated in feature space by a simple linear model. For example, when it is desired to predict whether someone will purchase a product based on many types of data such as their shopping history, surfing behavior, geographical location, etc., these data may be interrelated in a complex and non-linear manner such that the linear model cannot be accurately classified. In the medical field, predicting the risk of a disease based on patient genetic, lifestyle, disease history, etc. data also often involves complex nonlinear relationships. Complex data classification is widely used in a variety of fields, such as bioinformatics, financial risk management, medical diagnostics, and the like. By effectively classifying the complex data, the accuracy and efficiency of decision making can be improved.
Researchers have proposed a data classification method based on quantum support vector machines, which improves computational efficiency by exploiting parallelism and interferometry of quantum computation. The specific method for classifying the data based on the quantum support vector machine is that the quantum support vector machine encodes the original data onto a quantum state, and then processes the data through a quantum gate to obtain inner products among data samples, wherein the inner products can be regarded as inner products in a high-dimensional (even infinite-dimensional) feature space, namely a so-called quantum core. And finally, determining a classification hyperplane through a measurement result, thereby realizing data classification. The method can realize exponential acceleration when the feature quantity and the sample quantity are large.
However, although quantum computing may map data to a high-dimensional feature space, this mapping is linear, and for some data that have complex nonlinear relationships to each other, such linear mapping may not effectively separate the data, resulting in inaccurate classification results when the quantum support vector machine-based data classification method is processing complex data classifications.
Disclosure of Invention
In order to solve the problems, the invention provides a data classification method of a quantum support vector machine based on a variable component sub-core.
The method comprises the following steps:
step one, data to be classified are formed into a data set,/>Representing the number of data to be classified, +.>Representing the dimension of each piece of data to be classified, +.>Representation->Space of real number>Indicate->The data to be classified; data set +.>Stored in a quantum data structure;
step two, the data set is collectedBy mapping function->Mapping to obtain a mapping dataset +.>First->Strip data to be classified->Mapping to obtain mapping data to be classified>Mapping function->Is +.>;
Step three, preparing quantum characteristic codes of mapping data setsMapping data set +.>Performing quantum coding;
step four, according to the kernel function matrixDefining a nuclear mapping equation describing the mapping space data separation hyperplane parameters, and solving the nuclear mapping equation through a quantum linear solver to obtain a quantum state representing the mapping space data separation hyperplane parameters; kernel function matrix->Is->The value of each element in the kernel function matrix is the inner product of the mapping data to be classified represented by the row of the element and the mapping data to be classified represented by the column of the element;
fifthly, based on quantum states representing the mapping space data separation hyperplane parameters, mapping the data set by using a quantum support vector machineClassifying;
step six, constructing a loss function, and updating a mapping function based on the loss functionIs to be optimized for the parameter vector->Making the loss function value smaller than the preset value to obtain an optimized mapping function +.>;
Step seven, based on the optimized mapping functionMapping the data to be classified to obtain mapped data, and classifying the mapped data by using a quantum support vector machine.
Further, the mapping function in the second stepThe method comprises the following steps:
;
wherein,representing a mapping function +.>Is>Parameters to be optimized, < >>Representing a mapping function +.>The order of the non-linearity is such that,representing tensor product operations.
Further, mapping data set quantum characteristic code in step threeThe preparation process of (2) comprises:
according to quantum amplitude coding, all mapping data to be classifiedIs encoded into the data set quantum code +.>In (a):
;
wherein,representing two norms>Representing dirac symbols;
first, theMapping data to be classified->Quantum encoding of data to be classified->The method comprises the following steps:
;
wherein,indicate->Strip data to be classified->Is>Data of individual dimensionsMapping the obtained data;
preparation of quantum codes of each data to be classified based on quantum superposition parallel executionIs subjected to the operation of obtaining the mapping dataset quantum code +.>:
;
Quantum encoding based on mapping datasetComputation of the Quantum Property encoding of the mapping dataset>:
;
Wherein,quantum encoding representing mapping data sets>Corresponding density matrix, < >>Representing a trace operation on the density matrix, summing and eliminating information from the second subsystem.
Further, in the fourth step, the kernel mapping equation is:
;
;
;
wherein,representation is based on a kernel function matrix->Constructing a data association matrix->Representing a superparameter,/->Representing matrix transpose calculation, +.>Representing an identity matrix>Separating the normal vector of the hyperplane for mapping the spatial data, < >>Representing the intercept of the mapping spatial data separation hyperplane, < >>Is real number, < >>Representing mapping spatial data separation hyperplane parameters +.>Representation->Space of real number>For data category label->Representing a full vector.
Further, in the fourth step, the quantum state representing the mapped spatial data separation hyperplane parameter is obtained, specifically, representing the mapped spatial data separation hyperplane parameterQuantum state->The method comprises the following steps:
;
wherein,representing two norms>Indicate->Strip data to be classified->Weights separating hyperplane parameters for computation mapped spatial data, +.>Representing dirac symbols.
Further, the constructing the loss function in the step six specifically refers to obtaining the mapping dataset based on the hadamard testCalculating the mapping data set of the quantum support vector machine>Accuracy of classification->Definition description accuracy->And mapping function->Is to be optimized for the parameter vector->A loss function of the relationship.
Further, in step six, the mapping function is updated based on the loss functionIs to be optimized for the parameter vector->Specifically, it refers to:
calculating a loss function versus mapping functionIs to be optimized for the parameter vector->Is a gradient of (2); updating the mapping function +_with a predetermined learning rate according to the opposite direction of the gradient>Is to be optimized for the parameter vector->So that the loss function value gradually decreases.
One or more technical solutions provided in the embodiments of the present application at least have the following technical effects or advantages:
according to the invention, the variable component nonlinear kernel function framework is constructed, and the variable component nonlinear kernel function is used for mapping the data to be classified into the high-dimensional mapping space, so that the data to be classified can be separated by using a hyperplane in the mapping space, and the quantum support vector machine can process more complex data distribution, and the classification accuracy is improved. In addition, the invention also introduces a parameter optimization process, and the variable component nonlinear kernel function is more suitable for data distribution by optimizing parameters in the variable component nonlinear kernel function, so that the classification accuracy is further improved.
Drawings
Fig. 1 is a flowchart of a data classification method of a quantum support vector machine based on a variable component kernel according to an embodiment of the present invention.
Detailed Description
The present invention will be described in detail below with reference to the drawings and detailed embodiments, and before the technical solutions of the embodiments of the present invention are described in detail, the terms and terms involved will be explained, and in the present specification, the components with the same names or the same reference numerals represent similar or identical structures, and are only limited for illustrative purposes.
The data classification method based on the variable component sub-kernel comprises the steps of firstly constructing a variable component sub-nonlinear kernel function, namely a mapping function, mapping data to be classified into a mapping space through the mapping function, so that the mapped data can be separated by a hyperplane in the mapping space, and then dividing the mapped data through the quantum support vector machine to obtain a classification result. In order to construct a more proper mapping function, the invention introduces a parameter optimization process, and optimizes parameters in the mapping function by using a classification result of a quantum support vector machine, thereby better realizing classification tasks.
Fig. 1 shows specific steps of a data classification method of a quantum support vector machine based on a variable component kernel. The invention is explained below in connection with a specific embodiment. Assessing the risk of a person applying for a loan requires combining various personal data, such as the person's credit history, income, occupation, age, etc. The data classification method of the quantum support vector machine based on the variable component kernels is used for classifying loan applicants into two categories of loan applicants with high risk and loan applicants with low risk based on complex personal data.
1. Preserving data to be classified using a quantum data structure
Taking the collected personal data of loan applicant as data to be classified, wherein all the data to be classified form a data set,/>Representing data set +.>The number of data to be classified, +.>Representing data set +.>Dimension of each piece of data to be classified +.>Representation->Space of real number>Indicate->Data to be classified of the individual loan applicant, data to be classified +.>The corresponding data category label is->In this embodiment, <' > a->The label representing the data category is low risk, +.>The representative data category labels are high risk.
Data setStored in a quantum data structure. The present invention is not described in detail herein, taking the quantum data structure to store multidimensional data as common knowledge in the art.
2. Data mapping
Data setIn the case of a comparatively complex, the data set +.>The data to be classified in the data set can not be linearly separated, and the invention is used for the data setThe data to be classified in the data storage unit is subjected to nonlinear mapping, and the data to be classified is mapped to a mapping space with higher dimension, so that the mapped data to be classified can be separated by a hyperplane, and the hyperplane is defined as a mapping space data separation hyperplane.
Definition of the first embodimentData to be classified of individual loan applicant->The mapped data is the mapped data to be classified +.>:
;
Wherein,representing data set +.>Mapping function of mapping data to be classified into mapping space. When mapping function +.>In the case of a linear function, the data set cannot be hyperplane +.>Mapped data set ∈>And (5) separating. The invention maps the function->Defined as a nonlinear function comprising parameters to be optimized:
;
wherein,representing a mapping function +.>Is>Parameters to be optimized, < >>Representing a mapping function +.>To be optimized, parameter vector->Representing a mapping function +.>Nonlinear order, ++>Representing tensor product operations.
3. Preparation of mapping dataset Quantum states
To advantage ofMapping data sets using quantum technologyProcessing is carried out, and the mapping data set quantum characteristic code is needed to be prepared>Mapping data set quantum property coding>I.e. mapping data set +.>Is a quantum state of (c). Preparation of mapping dataset Quantum Property encoding +.>Can be expressed as:
;
wherein,represents +.>A kernel function matrix of dimension, wherein the value of each element in the kernel function matrix is the inner product of the mapping data to be classified represented by the row of the element and the mapping data to be classified represented by the column of the element, and the values of the elements in the kernel function matrix are->The trace operation is represented.
The construction of a kernel function matrix for data is common knowledge in the art, and the present invention will not be described in detail.
Mapping data set quantum property encodingsThe preparation process of (2) is as follows:
according to quantum amplitude coding, all mapping data to be classifiedIs encoded into the data set quantum code +.>In (a):
;
wherein,representing two norms>Representing dirac symbols.
In addition, the firstMapping data to be classified of individual loan applicant +.>Quantum encoding of data to be classified->The method comprises the following steps:
;
wherein,indicate->Data to be classified of individual loan applicant->Is>And mapping the data of each dimension to obtain data.
Preparation of quantum codes of each data to be classified based on quantum superposition parallel executionIs subjected to the operation of obtaining the mapping dataset quantum code +.>:
;
Quantum encoding based on mapping datasetComputation of the Quantum Property encoding of the mapping dataset>:
;
Wherein,quantum encoding representing mapping data sets>Corresponding density matrix, < >>Representing a trace operation on the density matrix, summing and eliminating information from the second subsystem.
4. Computing mapping spatial data separation hyperplane
Quantum character encoding based on mapping data setMapping data set +.>Encoded into quantum space and then calculated by quantumThe technique calculates a mapping spatial data separation hyperplane. Specifically, the invention uses a quantum support vector machine to find the mapping space data separation hyperplane, and the quantum support vector machine can prepare a quantum state representing the parameter of the mapping space data separation hyperplane, which is specifically shown as follows:
according to a kernel function matrixA system of linear equations describing the mapping space data separation hyperplane parameters is defined, which will be referred to herein as the nuclear mapping equation, which is expressed as:
;
;
;
wherein,representation is based on a kernel function matrix->Constructing a data association matrix->Representing a superparameter,/->Representing matrix transpose calculation, +.>Representing an identity matrix>Separating the normal vector of the hyperplane for mapping the spatial data, < >>Representing the intercept of the mapping spatial data separation hyperplane, < >>Is real number, < >>Representing mapping spatial data separation hyperplane parameters +.>Representation->Space of real number>Is->Data class label of individual loan applicant, +.>Representing a full vector.
Solving the nuclear mapping equation by a quantum linear solver to obtain the data separation hyperplane parameter representing the mapping spaceQuantum state->:
;
Wherein,representing two norms>Indicate->Data to be classified of individual loan applicant->The weights of the hyperplane are separated for the computed mapped spatial data.
A quantum linear solver such as a haro-Ha Xidi m-laud (HHL) solver or a quantum linear solver based on a quantum signal processing algorithm.
Solving the nuclear mapping equation based on the quantum linear solver is common knowledge in the art, and the invention is not repeated.
5. Loss function construction and parameter optimization
In the quantum support vector machine, obtaining the data separation hyperplane parameter representing the mapping spaceQuantum state->After that, can use->For mapping data set->The classification is performed, the classification result can be obtained through a Hadamard (Hadamard) test algorithm, and the classification result obtained based on the Hadamard test algorithm is common knowledge in the field, and the invention is not repeated.
Calculating the accuracy of classification according to the classification result. Accuracy->The value of (2) reflects the quantum support vector machine pair mapping dataset +.>Classification ability, accuracy->The larger the value of (c) is, the better the classification effect of the quantum support vector machine is.
Next by accuracy rateThe loss function is defined, and the loss function of the quantum support vector machine is defined based on accuracy, which is common knowledge in the art, and the invention is not repeated. Loss function and mapping function->Is to be optimized for the parameter vector->In relation, an optimal parameter vector to be optimized can be obtained by optimizing the loss function>。
Specifically, the loss function is optimized by a gradient descent method, and firstly, the loss function is calculated to optimize the parameter vectorIs reflected when changing the parameter vector to be optimized +.>The direction and rate of change of the loss function. Then, the parameter vector to be optimized is updated according to the opposite direction of the gradient at a preset learning rate>Gradually reducing the loss function value until the loss function value is smaller than a preset value, obtaining an optimized mapping function +.>。
6. Data classification
Obtaining an optimized mapping functionAfter that, based on the optimized mapping function +.>After the personal data of the loan applicant to be classified is mapped, the classification is performed by using a quantum support vector machine.
The above embodiments are merely illustrative of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention, and various modifications and improvements made by those skilled in the art to the technical solution of the present invention should fall within the protection scope defined by the claims of the present invention without departing from the design spirit of the present invention.
Claims (6)
1. The data classification method of the quantum support vector machine based on the variable component sub-core is characterized by comprising the following steps of:
step one, data to be classified are formed into a data set,/>Representing the number of data to be classified, +.>Representing the dimension of each piece of data to be classified, +.>Representation->Space of real number>Indicate->Stripe data to be classifiedThe method comprises the steps of carrying out a first treatment on the surface of the Data set +.>Stored in a quantum data structure;
step two, the data set is collectedBy mapping function->Mapping to obtain a mapping dataset +.>First->Strip data to be classified->Mapping to obtain mapping data to be classified>Mapping function->Is +.>The method comprises the steps of carrying out a first treatment on the surface of the Said mapping function->The method comprises the following steps:
;
wherein,representing a mapping function +.>Is>Parameters to be optimized, < >>Representing a mapping function +.>Nonlinear order, ++>Representing tensor product operations;
step three, preparing quantum characteristic codes of mapping data setsMapping data set +.>Performing quantum coding;
step four, according to the kernel function matrixDefining a nuclear mapping equation describing the mapping space data separation hyperplane parameters, and solving the nuclear mapping equation through a quantum linear solver to obtain a quantum state representing the mapping space data separation hyperplane parameters; kernel function matrix->Is->The value of each element in the kernel function matrix is the inner product of the mapping data to be classified represented by the row of the element and the mapping data to be classified represented by the column of the element;
step five, separating quantum state of hyperplane parameter based on representation mapping space dataMapping data sets using a quantum support vector machine pairClassifying;
step six, constructing a loss function, and updating a mapping function based on the loss functionIs to be optimized for the parameter vector->Making the loss function value smaller than the preset value to obtain an optimized mapping function +.>;
Step seven, based on the optimized mapping functionMapping the data to be classified to obtain mapped data, and classifying the mapped data by using a quantum support vector machine.
2. The method for classifying data based on variable component kernel quantum support vector machine according to claim 1, wherein the mapping data set quantum characteristic code in step threeThe preparation process of (2) comprises:
according to quantum amplitude coding, all mapping data to be classifiedIs encoded into the data set quantum code +.>In (a):
;
wherein,representing two norms>Representing dirac symbols;
first, theMapping data to be classified->Quantum encoding of data to be classified->The method comprises the following steps:
;
wherein,indicate->Strip data to be classified->Is>Data obtained after mapping the data of each dimension;
preparation of quantum codes of each data to be classified based on quantum superposition parallel executionIs subjected to the operation of obtaining the mapping dataset quantum code +.>:
;
Quantum encoding based on mapping datasetComputation of the Quantum Property encoding of the mapping dataset>:
;
Wherein,quantum encoding representing mapping data sets>Corresponding density matrix, < >>Representing a trace operation on the density matrix, summing and eliminating information from the second subsystem.
3. The method for classifying data based on a variable component kernel quantum support vector machine according to claim 1, wherein in the fourth step, the kernel mapping equation is:
;
;
;
wherein,representation is based on a kernel function matrix->Constructing a data association matrix->Representing a superparameter,/->Representing matrix transpose calculation, +.>Representing an identity matrix>Separating the normal vector of the hyperplane for mapping the spatial data, < >>Representing the intercept of the mapping spatial data separation hyperplane, < >>Is real number, < >>Representing mapping spatial data separation hyperplane parameters +.>Representation->Space of real number>For data category label->Representing a full vector.
4. The method for classifying data by using variable component kernel-based quantum support vector machine according to claim 3, wherein in the fourth step, the obtained quantum state representing the mapped spatial data separation hyperplane parameter is specifically represented by the mapped spatial data separation hyperplane parameterQuantum state->The method comprises the following steps:
;
wherein,representing two norms>Indicate->Strip data to be classified->Weights separating hyperplane parameters for computation mapped spatial data, +.>Representing dirac symbols.
5. A according to claim 1The data classification method of the quantum support vector machine based on the variable component sub-core is characterized in that the construction of the loss function in the step six specifically refers to obtaining a mapping data set based on Hadamard testCalculating the mapping data set of the quantum support vector machine>Accuracy of classification->Definition description accuracy->And mapping function->Is to be optimized for the parameter vector->A loss function of the relationship.
6. The method for classifying data based on variable component kernel quantum support vector machine according to claim 1, wherein in step six, the mapping function is updated based on a loss functionIs to be optimized for the parameter vector->Specifically, it refers to:
calculating a loss function versus mapping functionIs to be optimized for the parameter vector->Is a gradient of (2); updating the mapping function +_with a predetermined learning rate according to the opposite direction of the gradient>Is to be optimized for the parameter vector->So that the loss function value gradually decreases.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410071683.3A CN117591947B (en) | 2024-01-18 | 2024-01-18 | Data classification method of quantum support vector machine based on variable component sub-core |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410071683.3A CN117591947B (en) | 2024-01-18 | 2024-01-18 | Data classification method of quantum support vector machine based on variable component sub-core |
Publications (2)
Publication Number | Publication Date |
---|---|
CN117591947A CN117591947A (en) | 2024-02-23 |
CN117591947B true CN117591947B (en) | 2024-04-09 |
Family
ID=89918685
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410071683.3A Active CN117591947B (en) | 2024-01-18 | 2024-01-18 | Data classification method of quantum support vector machine based on variable component sub-core |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117591947B (en) |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106053067A (en) * | 2016-05-24 | 2016-10-26 | 广东石油化工学院 | Bearing fault diagnosis method based on quantum genetic algorithm optimized support vector machine |
CN108595555A (en) * | 2018-04-11 | 2018-09-28 | 西安电子科技大学 | The image search method returned based on semi-supervised tensor subspace |
CN109194236A (en) * | 2018-09-26 | 2019-01-11 | 河南科技大学 | Based on the induction-type bearingless motor of LS-SVM without radial displacement transducer control system |
CN111369003A (en) * | 2020-03-31 | 2020-07-03 | 合肥本源量子计算科技有限责任公司 | Method and device for determining fidelity of quantum bit reading signal |
CN112686328A (en) * | 2021-01-06 | 2021-04-20 | 成都信息工程大学 | Data classification system and method based on quantum fuzzy information |
CN113661501A (en) * | 2019-04-03 | 2021-11-16 | 国际商业机器公司 | Quantum feature nuclear alignment |
CN114186620A (en) * | 2021-11-30 | 2022-03-15 | 广东电网有限责任公司 | Multi-dimensional training method and device for support vector machine |
CN114358319A (en) * | 2022-03-22 | 2022-04-15 | 合肥本源量子计算科技有限责任公司 | Machine learning framework-based classification method and related device |
WO2023093857A1 (en) * | 2021-11-26 | 2023-06-01 | 本源量子计算科技(合肥)股份有限公司 | Method and apparatus for solving system of nonlinear equations on basis of quantum circuit, and storage medium |
EP4242934A1 (en) * | 2022-03-07 | 2023-09-13 | Qu&CO R&D B.V. | Quantum-kernel-based regression |
CN117010519A (en) * | 2023-08-11 | 2023-11-07 | 中国银行股份有限公司 | Data processing method of quantum support vector machine, storage medium and related equipment |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3646250A1 (en) * | 2017-05-30 | 2020-05-06 | GTN Ltd | Tensor network machine learning system |
US11295223B2 (en) * | 2018-06-12 | 2022-04-05 | International Business Machines Corporation | Quantum feature kernel estimation using an alternating two layer quantum circuit |
US11562279B2 (en) * | 2019-06-18 | 2023-01-24 | Xanadu Quantum Technologies Inc. | Apparatus and methods for quantum computing and machine learning |
-
2024
- 2024-01-18 CN CN202410071683.3A patent/CN117591947B/en active Active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106053067A (en) * | 2016-05-24 | 2016-10-26 | 广东石油化工学院 | Bearing fault diagnosis method based on quantum genetic algorithm optimized support vector machine |
CN108595555A (en) * | 2018-04-11 | 2018-09-28 | 西安电子科技大学 | The image search method returned based on semi-supervised tensor subspace |
CN109194236A (en) * | 2018-09-26 | 2019-01-11 | 河南科技大学 | Based on the induction-type bearingless motor of LS-SVM without radial displacement transducer control system |
CN113661501A (en) * | 2019-04-03 | 2021-11-16 | 国际商业机器公司 | Quantum feature nuclear alignment |
CN111369003A (en) * | 2020-03-31 | 2020-07-03 | 合肥本源量子计算科技有限责任公司 | Method and device for determining fidelity of quantum bit reading signal |
CN112686328A (en) * | 2021-01-06 | 2021-04-20 | 成都信息工程大学 | Data classification system and method based on quantum fuzzy information |
WO2023093857A1 (en) * | 2021-11-26 | 2023-06-01 | 本源量子计算科技(合肥)股份有限公司 | Method and apparatus for solving system of nonlinear equations on basis of quantum circuit, and storage medium |
CN114186620A (en) * | 2021-11-30 | 2022-03-15 | 广东电网有限责任公司 | Multi-dimensional training method and device for support vector machine |
EP4242934A1 (en) * | 2022-03-07 | 2023-09-13 | Qu&CO R&D B.V. | Quantum-kernel-based regression |
CN114358319A (en) * | 2022-03-22 | 2022-04-15 | 合肥本源量子计算科技有限责任公司 | Machine learning framework-based classification method and related device |
CN117010519A (en) * | 2023-08-11 | 2023-11-07 | 中国银行股份有限公司 | Data processing method of quantum support vector machine, storage medium and related equipment |
Non-Patent Citations (4)
Title |
---|
Quantum support vector machine for big data classification;Patrick Rebentrost et al.;《https://arxiv.org/abs/1307.0471》;20140710;全文 * |
基于参数化量子电路的量子卷积神经网络模型及应用;郑瑾 等;《控制理论与应用》;20211130;第38卷(第11期);1772-1784 * |
若干求解非线性方程的量子算法研究;薛程;《中国博士学位论文全文数据库基础科学辑》;20230315(第3期);全文 * |
量子支持向量机算法研究;崔振宇;《中国优秀硕士学位论文全文数据库信息科技辑》;20220815(第8期);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN117591947A (en) | 2024-02-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112070125A (en) | Prediction method of unbalanced data set based on isolated forest learning | |
CN110163261A (en) | Unbalanced data disaggregated model training method, device, equipment and storage medium | |
Borcard et al. | Unconstrained ordination | |
EP3792840A1 (en) | Neural network method and apparatus | |
CN111352965B (en) | Training method of sequence mining model, and processing method and equipment of sequence data | |
Cui et al. | Internet financing credit risk evaluation using multiple structural interacting elastic net feature selection | |
CN111382930B (en) | Time sequence data-oriented risk prediction method and system | |
WO2016183391A1 (en) | System, method and computer-accessible medium for making a prediction from market data | |
Gan et al. | Data clustering with actuarial applications | |
Mohammadi et al. | Improving linear discriminant analysis with artificial immune system-based evolutionary algorithms | |
Tang et al. | Gene selection for microarray data classification via subspace learning and manifold regularization | |
Tsai et al. | A fast particle swarm optimization for clustering | |
CN113033090A (en) | Push model training method, data push device and storage medium | |
CN114154557A (en) | Cancer tissue classification method, apparatus, electronic device, and storage medium | |
Xie et al. | Unsupervised feature extraction with convolutional autoencoder with application to daily stock market prediction | |
Li et al. | An ensemble clustering framework based on hierarchical clustering ensemble selection and clusters clustering | |
He et al. | Self-Adaptive bagging approach to credit rating | |
Chen | A hybrid model for business failure prediction-utilization of particle swarm optimization and support vector machines | |
CN117591947B (en) | Data classification method of quantum support vector machine based on variable component sub-core | |
CN117273157A (en) | Quantum core method, classification method, data coding method, related system and device | |
Somogyi et al. | Performance evaluation of machine learning models | |
CN114549174A (en) | User behavior prediction method and device, computer equipment and storage medium | |
El Mouden et al. | A graph-based methodology for tracking covid-19 in time series datasets | |
Jia et al. | Auto weighted robust dual graph nonnegative matrix factorization for multiview clustering | |
CN115063143A (en) | Account data processing method and device, computer equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |