CN108664992B - Classification method and device based on genetic optimization and kernel extreme learning machine - Google Patents
Classification method and device based on genetic optimization and kernel extreme learning machine Download PDFInfo
- Publication number
- CN108664992B CN108664992B CN201810295090.XA CN201810295090A CN108664992B CN 108664992 B CN108664992 B CN 108664992B CN 201810295090 A CN201810295090 A CN 201810295090A CN 108664992 B CN108664992 B CN 108664992B
- Authority
- CN
- China
- Prior art keywords
- kernel
- learning machine
- kernel function
- population
- pictures
- 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
Images
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
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
Abstract
The invention discloses a classification method based on genetic optimization and a kernel limit learning machine, which comprises the following steps: inputting the pictures in the sample picture library into a recursive least square method kernel limit learning machine; obtaining the classification precision of classifying the pictures in the sample picture library under a plurality of groups of different normalization coefficients and kernel function parameters, wherein the classification result comprises a preset number of user preference levels; optimizing the normalization coefficient and the kernel function parameter based on a genetic algorithm to obtain the optimal normalization coefficient and kernel function parameter; inputting the optimal normalization coefficient and the value of the kernel function parameter into a recursive least square method kernel limit learning machine so as to classify the pictures in the picture library to be classified; and recommending the pictures to the user from high to low according to the corresponding user favorite levels. In the method, the classification parameters in the recursive least square method kernel extreme learning machine are optimal classification parameters obtained after genetic algorithm optimization, the classification precision is improved, and the picture recommendation accuracy is further improved.
Description
Technical Field
The invention relates to the technical field of data classification, in particular to a classification method based on genetic optimization and a nuclear extreme learning machine, and further relates to a classification device, equipment and a computer readable storage medium based on the genetic optimization and the nuclear extreme learning machine.
Background
The picture recommendation application or the picture recommendation webpage can recommend pictures to the user according to the preference of the user, wherein data classification is important for accurate recommendation of the pictures. The existing data classification methods include a neural network method, a support vector machine method, a decision tree classification method, a Bayesian network method, a recursive least square method kernel limit learning machine and the like, wherein the recursive least square method kernel limit learning machine is a commonly used data classification means, but when the recursive least square method is used, a user needs to select classification parameters according to test results after repeated tests, when the data volume is large, the method easily causes unreliable selection of the classification parameters due to subjective judgment of the user, causes low classification precision, and further causes reduced picture recommendation accuracy, namely, the probability that pictures recommended to the user are loved by the user is reduced.
In summary, how to provide a scheme capable of improving the accuracy of image recommendation is a problem to be urgently solved by those skilled in the art.
Disclosure of Invention
The invention aims to provide a classification method, a classification device, a classification equipment and a computer readable storage medium based on genetic optimization and a kernel limit learning machine, which can improve the accuracy of picture recommendation.
In order to achieve the above purpose, the invention provides the following technical scheme:
a classification method based on genetic optimization and a kernel-limit learning machine comprises the following steps:
inputting the pictures in the sample picture library into a recursive least square method kernel limit learning machine;
obtaining classification precision of the recursive least square method kernel limit learning machine for classifying the pictures in the sample picture library under a plurality of groups of different normalization coefficients and kernel function parameters, wherein one group of normalization coefficients and kernel function parameters correspond to one classification precision, and classification results of the recursive least square method kernel limit learning machine for the pictures in the sample picture library comprise a preset number of user preference levels;
optimizing the normalization coefficient and the kernel function parameter based on a genetic algorithm to obtain the optimal normalization coefficient and the optimal kernel function parameter, wherein the classification precision corresponding to any group of normalization coefficients and the kernel function parameter is a fitness value in the genetic algorithm;
inputting the optimal normalization coefficient and the value of the kernel function parameter into the recursive least square method kernel limit learning machine so as to classify the pictures in the picture library to be classified;
and recommending the pictures to the user from high to low according to the corresponding user favorite levels.
Preferably, the optimizing the normalization coefficient and the kernel function parameter based on the genetic algorithm to obtain the optimal normalization coefficient and kernel function parameter includes:
taking the multiple groups of different normalization coefficients and kernel function parameters as individuals of a population in a genetic algorithm, wherein all the individuals form a first population;
randomly generating a second population having a preset number of individuals from the first population;
removing individuals with fitness values smaller than a preset first fitness value from the second population;
performing cross calculation and variation calculation on the first population after the first population is removed;
adding 1 to the current genetic algebra;
and judging whether the current genetic algebra reaches a preset genetic algebra, if so, stopping heredity, finding out the individual with the maximum fitness value in the current second population, determining the normalization coefficient and the kernel function parameter corresponding to the individual as the optimal normalization coefficient and kernel function parameter values, if not, returning to execute the step of eliminating the individual with the fitness value smaller than the preset first fitness value from the second population until the current genetic algebra is judged to be equal to the preset genetic algebra.
Preferably, the kernel function in the least square method kernel limit learning machine is a quadratic rational kernel function.
Preferably, before inputting the pictures in the sample picture library into the recursive least square method kernel-limit learning machine, the method further includes:
and carrying out normalization processing on the picture data in the sample picture library.
A classification apparatus based on genetic optimization and a kernel-limit learning machine, comprising:
a first input unit to: inputting the pictures in the sample picture library into a recursive least square method kernel limit learning machine;
an acquisition unit configured to: obtaining classification precision of the recursive least square method kernel limit learning machine for classifying the pictures in the sample picture library under a plurality of groups of different normalization coefficients and kernel function parameters, wherein one group of normalization coefficients and kernel function parameters correspond to one classification precision, and classification results of the recursive least square method kernel limit learning machine for the pictures in the sample picture library comprise a preset number of user preference levels;
an optimization unit for: optimizing the normalization coefficient and the kernel function parameter based on a genetic algorithm to obtain the optimal normalization coefficient and the optimal kernel function parameter, wherein the classification precision corresponding to any group of normalization coefficients and the kernel function parameter is a fitness value in the genetic algorithm;
a second input unit for: inputting the optimal normalization coefficient and the value of the kernel function parameter into the recursive least square method kernel limit learning machine so as to classify the pictures in the picture library to be classified;
a recommendation unit to: and recommending the pictures to the user from high to low according to the corresponding user favorite levels.
Preferably, the optimization unit includes:
a generating unit configured to: taking the multiple groups of different normalization coefficients and kernel function parameters as individuals of a population in a genetic algorithm, wherein all the individuals form a first population; randomly generating a second population having a preset number of individuals from the first population;
a selection unit for: removing individuals with fitness values smaller than a preset first fitness value from the second population;
a crossover and mutation unit to: performing cross calculation and variation calculation on the first population after the first population is removed;
a counting unit for: adding 1 to the current genetic algebra;
a determination unit configured to: and judging whether the current genetic algebra reaches a preset genetic algebra, if so, stopping heredity, finding out the individual with the maximum fitness value in the current second population, determining the normalization coefficient and the kernel function parameter corresponding to the individual as the optimal normalization coefficient and kernel function parameter values, if not, returning to execute the step of eliminating the individual with the fitness value smaller than the preset first fitness value from the second population until the current genetic algebra is judged to be equal to the preset genetic algebra.
Preferably, the kernel function in the least square method kernel limit learning machine is a quadratic rational kernel function.
Preferably, the method further comprises the following steps:
a normalization processing unit configured to: before the first input unit inputs the pictures in the sample picture library into the recursive least square method kernel limit learning machine, the picture data in the sample picture library is normalized.
A classification apparatus based on genetic optimization and a kernel-limit learning machine, comprising:
a memory to: storing a computer program;
a processor to: the computer program when executed implements the steps of a genetic optimization and kernel-extreme learning machine based classification method as defined in any one of the above.
A computer-readable storage medium having stored thereon a computer program which, when executed by a processor, carries out the steps of a genetic optimization and kernel-limit learning machine based classification method according to any one of the preceding claims.
The invention provides a classification method based on genetic optimization and a kernel limit learning machine, which comprises the following steps: inputting the pictures in the sample picture library into a recursive least square method kernel limit learning machine; obtaining classification precision of the recursive least square method kernel limit learning machine for classifying the pictures in the sample picture library under a plurality of groups of different normalization coefficients and kernel function parameters, wherein one group of normalization coefficients and kernel function parameters correspond to one classification precision, and classification results of the recursive least square method kernel limit learning machine for the pictures in the sample picture library comprise a preset number of user preference levels; optimizing the normalization coefficient and the kernel function parameter based on a genetic algorithm to obtain the optimal normalization coefficient and the optimal kernel function parameter, wherein the classification precision corresponding to any group of normalization coefficients and the kernel function parameter is a fitness value in the genetic algorithm; inputting the optimal normalization coefficient and the value of the kernel function parameter into the recursive least square method kernel limit learning machine so as to classify the pictures in the picture library to be classified; and recommending the pictures to the user from high to low according to the corresponding user favorite levels. In the method, the classification parameters in the recursive least square method kernel extreme learning machine are optimal classification parameters obtained after genetic algorithm optimization, the classification precision is improved, the picture recommendation accuracy is further improved, the user experience is further improved, and the user viscosity is further increased.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
FIG. 1 is a flow chart of a classification method based on genetic optimization and a kernel-limit learning machine according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a classification device based on genetic optimization and a kernel limit learning machine according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, a classification method based on genetic optimization and a kernel-limit learning machine according to an embodiment of the present invention is shown, which may include:
step S11: and inputting the pictures in the sample picture library into a recursive least square method kernel limit learning machine.
And the pictures in the sample picture library are used for training a recursive least square method kernel limit learning machine.
Step S12: and obtaining the classification precision of the recursive least square method kernel limit learning machine for classifying the pictures in the sample picture library under a plurality of groups of different normalization coefficients and kernel function parameters, wherein one group of normalization coefficients and kernel function parameters correspond to one classification precision, and the classification result of the recursive least square method kernel limit learning machine on the pictures in the sample picture library comprises a preset number of user preference levels.
Firstly, determining the variation range of a normalization coefficient and a kernel function parameter, then randomly generating a plurality of groups of normalization coefficients and kernel function parameters with specific values according to the determined variation range, respectively substituting different groups of normalization coefficients and kernel function parameters into a recursive least square method kernel limit learning machine to correspondingly obtain different models of the recursive least square method kernel limit learning machine, and then respectively classifying pictures in a sample picture library by using different models. The standard user preference level of the pictures input into the recursive least square method kernel-based extreme learning machine is predetermined, so that the classification precision of each model can be obtained by comparing the classification result of the different models of the recursive least square method kernel-based extreme learning machine on the pictures in the sample picture library with the predetermined standard user preference level. The user's preference level can be divided according to the actual situation, for example, it can be divided into: like, general, dislike, etc.
Step S13: and optimizing the normalization coefficients and the kernel function parameters based on the genetic algorithm to obtain the optimal normalization coefficients and the optimal kernel function parameters, wherein the classification precision corresponding to any group of normalization coefficients and the kernel function parameters is a fitness value in the genetic algorithm.
The genetic algorithm is an efficient global optimization searching algorithm which is based on natural selection and genetic theory and combines the survival rule of the fittest in the biological evolution process with the random information exchange mechanism of chromosomes in the population. The optimal normalization coefficient and kernel function parameters can be determined through a genetic algorithm.
Step S14: and inputting the optimal normalization coefficient and the value of the kernel function parameter into a recursive least square method kernel limit learning machine so as to classify the pictures in the picture library to be classified.
After the optimal normalization coefficient and the value of the kernel function parameter are determined, the optimal value can be input into a recursive least square kernel limit learning machine to obtain an optimal recursive least square kernel limit learning machine model, the model is a model which is trained through pictures in a sample picture library, and then the model is used for classifying the pictures in the picture library to be classified.
Step S15: and recommending the pictures to the user from high to low according to the corresponding user favorite levels.
The higher the favorite level of the picture is, the more the user likes the picture, so that the user is recommended with the high favorite level first and then with the low favorite level.
In the method, the classification parameters in the recursive least square method kernel extreme learning machine are optimal classification parameters obtained after genetic algorithm optimization, the classification precision is improved, the picture recommendation accuracy is further improved, the user experience is further improved, and the user viscosity is further increased.
Preferably, optimizing the normalization coefficient and the kernel function parameter based on a genetic algorithm to obtain the optimal normalization coefficient and kernel function parameter includes:
taking a plurality of groups of different normalization coefficients and kernel function parameters as individuals of a population in a genetic algorithm, wherein all the individuals form a first population;
randomly generating a second population having a preset number of individuals from the first population;
removing individuals with fitness values smaller than a preset first fitness value from the second population;
performing cross calculation and variation calculation on the first population after the first population is removed;
adding 1 to the current genetic algebra;
and judging whether the current genetic algebra reaches a preset genetic algebra, if so, stopping heredity, finding out the individual with the maximum fitness value in the current second population, determining the normalization coefficient and the kernel function parameter corresponding to the individual as the optimal normalization coefficient and kernel function parameter values, if not, returning to execute the step of removing the individual with the fitness value smaller than the preset first fitness value from the second population until the current genetic algebra is judged to be equal to the preset genetic algebra.
The first population is composed of total sample data, and a second population is randomly generated from the first population, wherein the number of individuals in the second population is set according to actual conditions, such as: considering the size of the data volume to be classified, if the data volume to be classified is large, the number of individuals of the second population may be set to be larger, and otherwise, may be set to be smaller. In the actual implementation process, the individuals in the first population are usually encoded, the encoding mode can be set according to the actual situation, binary encoding is usually adopted, and the encoding length is set according to the actual data size. The cross calculation comprises single-point cross calculation, multi-point cross calculation, uniform cross, shuffling cross, reduction agent cross calculation and the like, and can be selected according to actual conditions. And each time the elimination, the cross calculation and the variation calculation in the process are completed, the heredity is completed, and the heredity result is that the individuals with the fitness value not less than the first fitness value are inherited and new and better individuals are generated. The target genetic algebra is preset, if the genetic algebra reaches a preset value, the inheritance can be stopped, otherwise, the inheritance is continued.
Preferably, the kernel function in the least square method kernel limit learning machine is a quadratic rational kernel function.
The secondary rational kernel function has wide action, can improve the classification speed of the least square method kernel extreme learning machine and reduce the time consumption.
Preferably, before inputting the pictures in the sample picture library into the recursive least square method kernel limit learning machine, the method further includes: and carrying out normalization processing on the picture data in the sample picture library.
After normalization processing, the range of data is between 0 and 1, and the processing speed of the recursive least square method kernel limit learning machine can be improved.
Referring to fig. 2, a classification apparatus based on genetic optimization and a kernel-limit learning machine according to an embodiment of the present invention is shown, which may include:
a first input unit 11 for: inputting the pictures in the sample picture library into a recursive least square method kernel limit learning machine;
an acquisition unit 12 configured to: obtaining classification precision of classifying the pictures in the sample picture library by the recursive least square method kernel limit learning machine under a plurality of groups of different normalization coefficients and kernel function parameters, wherein one group of normalization coefficients and kernel function parameters correspond to one classification precision, and classification results of the pictures in the sample picture library by the recursive least square method kernel limit learning machine comprise a preset number of user preference levels;
an optimization unit 13 for: optimizing normalization coefficients and kernel function parameters based on a genetic algorithm to obtain optimal normalization coefficients and kernel function parameters, wherein the classification precision corresponding to any group of normalization coefficients and kernel function parameters is a fitness value in the genetic algorithm;
a second input unit 14 for: inputting the optimal normalization coefficient and the value of the kernel function parameter into a recursive least square method kernel limit learning machine so as to classify the pictures in the picture library to be classified;
a recommendation unit 15 configured to: and recommending the pictures to the user from high to low according to the corresponding user favorite levels.
Preferably, the optimization unit 13 may include:
a generating unit configured to: taking a plurality of groups of different normalization coefficients and kernel function parameters as individuals of a population in a genetic algorithm, wherein all the individuals form a first population; randomly generating a second population having a preset number of individuals from the first population;
a selection unit for: removing individuals with fitness values smaller than a preset first fitness value from the second population;
a crossover and mutation unit to: performing cross calculation and variation calculation on the first population after the first population is removed;
a counting unit for: adding 1 to the current genetic algebra;
a determination unit configured to: and judging whether the current genetic algebra reaches a preset genetic algebra, if so, stopping heredity, finding out the individual with the maximum fitness value in the current second population, determining the normalization coefficient and the kernel function parameter corresponding to the individual as the optimal normalization coefficient and kernel function parameter values, if not, returning to execute the step of removing the individual with the fitness value smaller than the preset first fitness value from the second population until the current genetic algebra is judged to be equal to the preset genetic algebra.
Preferably, the kernel function in the least square method kernel limit learning machine is a quadratic rational kernel function.
Preferably, the classification device based on genetic optimization and a kernel limit learning machine according to an embodiment of the present invention may further include:
a normalization processing unit configured to: before the first input unit inputs the pictures in the sample picture library into the recursive least square method kernel limit learning machine, the picture data in the sample picture library is subjected to normalization processing.
For a description of a relevant part in the classification device based on genetic optimization and the kernel-based extreme learning machine provided by the embodiment of the present invention, please refer to a detailed description of a corresponding part in the classification method based on genetic optimization and the kernel-based extreme learning machine provided by the embodiment of the present invention, which is not repeated herein.
The embodiment of the invention also provides a classification device based on genetic optimization and a kernel limit learning machine, which comprises:
a memory to: storing a computer program;
a processor to: the computer program when executed implements the steps of a genetic optimization and kernel-extreme learning machine-based classification method as in one of the above.
The invention also provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of any one of the above-described methods of classification based on genetic optimization and a nuclear limit learning machine.
In addition, parts of the above technical solutions provided in the embodiments of the present invention that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail, so as to avoid redundant description.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
It is further noted that, in the present specification, relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
Claims (8)
1. A classification method based on genetic optimization and a kernel-limit learning machine is characterized by comprising the following steps:
inputting the pictures in the sample picture library into a recursive least square method kernel limit learning machine;
obtaining classification precision of the recursive least square method kernel limit learning machine for classifying the pictures in the sample picture library under multiple groups of different regularization coefficients and kernel function parameters, wherein one group of regularization coefficients and kernel function parameters correspond to one classification precision, and classification results of the recursive least square method kernel limit learning machine for the pictures in the sample picture library comprise preset number of user preference levels;
optimizing the regularization coefficients and the kernel function parameters based on a genetic algorithm to obtain optimal regularization coefficients and kernel function parameters, wherein the classification precision corresponding to any group of regularization coefficients and kernel function parameters is a fitness value in the genetic algorithm;
inputting the optimal regularization coefficient and the value of the kernel function parameter into the recursive least square method kernel limit learning machine so as to classify the pictures in the picture library to be classified;
recommending the pictures to the user from high to low according to the corresponding user favorite levels;
optimizing the regularization coefficient and the kernel function parameter based on a genetic algorithm to obtain the optimal regularization coefficient and the optimal kernel function parameter, wherein the method comprises the following steps:
taking a plurality of groups of different regularization coefficients and kernel function parameters as individuals of a population in a genetic algorithm, wherein all the individuals form a first population;
randomly generating a second population having a preset number of individuals from the first population;
removing individuals with fitness values smaller than a preset first fitness value from the second population;
performing cross calculation and variation calculation on the first population after the first population is removed;
adding 1 to the current genetic algebra;
and judging whether the current genetic algebra reaches a preset genetic algebra, if so, stopping heredity, finding out the individual with the maximum fitness value in the current second population, determining the regularization coefficient and the kernel function parameter corresponding to the individual as the optimal regularization coefficient and kernel function parameter values, if not, returning to execute the step of removing the individual with the fitness value smaller than the preset first fitness value from the second population until the current genetic algebra is judged to be equal to the preset genetic algebra.
2. The method of claim 1, wherein the kernel function in the least squares kernel-extreme learning machine is a quadratic rational kernel function.
3. The method of claim 2, wherein before inputting the pictures in the sample picture library into the recursive least squares kernel-limit learning machine, further comprising:
and carrying out normalization processing on the picture data in the sample picture library.
4. A classification apparatus based on genetic optimization and a kernel-limit learning machine, comprising:
a first input unit to: inputting the pictures in the sample picture library into a recursive least square method kernel limit learning machine;
an acquisition unit configured to: obtaining classification precision of the recursive least square method kernel limit learning machine for classifying the pictures in the sample picture library under multiple groups of different regularization coefficients and kernel function parameters, wherein one group of regularization coefficients and kernel function parameters correspond to one classification precision, and classification results of the recursive least square method kernel limit learning machine for the pictures in the sample picture library comprise preset number of user preference levels;
an optimization unit for: optimizing the regularization coefficients and the kernel function parameters based on a genetic algorithm to obtain optimal regularization coefficients and kernel function parameters, wherein the classification precision corresponding to any group of regularization coefficients and kernel function parameters is a fitness value in the genetic algorithm;
a second input unit for: inputting the optimal regularization coefficient and the value of the kernel function parameter into the recursive least square method kernel limit learning machine so as to classify the pictures in the picture library to be classified;
a recommendation unit to: recommending the pictures to the user from high to low according to the corresponding user favorite levels; the optimization unit includes:
a generating unit configured to: taking the different regularization coefficients and kernel function parameters as individuals of a population in a genetic algorithm, wherein all the individuals form a first population; randomly generating a second population having a preset number of individuals from the first population;
a selection unit for: removing individuals with fitness values smaller than a preset first fitness value from the second population;
a crossover and mutation unit to: performing cross calculation and variation calculation on the first population after the first population is removed;
a counting unit for: adding 1 to the current genetic algebra;
a determination unit configured to: and judging whether the current genetic algebra reaches a preset genetic algebra, if so, stopping heredity, finding out the individual with the maximum fitness value in the current second population, determining the regularization coefficient and the kernel function parameter corresponding to the individual as the optimal regularization coefficient and kernel function parameter value, if not, returning to execute the step of eliminating the individual with the fitness value smaller than the preset first fitness value from the second population until the current genetic algebra is judged to be equal to the preset genetic algebra.
5. The apparatus of claim 4, wherein the kernel function in the least squares kernel-extreme learning machine is a quadratic rational kernel function.
6. The apparatus of claim 5, further comprising:
a normalization processing unit configured to: before the first input unit inputs the pictures in the sample picture library into the recursive least square method kernel limit learning machine, the picture data in the sample picture library is normalized.
7. A classification device based on genetic optimization and a kernel-limit learning machine, comprising:
a memory to: storing a computer program;
a processor to: the computer program when executed implements the steps of a genetic optimization and kernel-extreme learning machine based classification method according to any one of claims 1 to 3.
8. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a computer program which, when being executed by a processor, carries out the steps of a genetic optimization and kernel-limit learning machine-based classification method according to any one of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810295090.XA CN108664992B (en) | 2018-03-30 | 2018-03-30 | Classification method and device based on genetic optimization and kernel extreme learning machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810295090.XA CN108664992B (en) | 2018-03-30 | 2018-03-30 | Classification method and device based on genetic optimization and kernel extreme learning machine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108664992A CN108664992A (en) | 2018-10-16 |
| CN108664992B true CN108664992B (en) | 2022-02-15 |
Family
ID=63783069
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810295090.XA Active CN108664992B (en) | 2018-03-30 | 2018-03-30 | Classification method and device based on genetic optimization and kernel extreme learning machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108664992B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103400190A (en) * | 2013-08-13 | 2013-11-20 | 浙江大学 | Integrated framework method for optimizing extremity learning machine by using genetic algorithm |
| CN104537382A (en) * | 2015-01-12 | 2015-04-22 | 杭州电子科技大学 | Electromyographic signal gait recognition method for optimizing support vector machine based on genetic algorithm |
| CN104657718A (en) * | 2015-02-13 | 2015-05-27 | 武汉工程大学 | Face recognition method based on face image feature extreme learning machine |
| CN104700078A (en) * | 2015-02-13 | 2015-06-10 | 武汉工程大学 | Scale-invariant feature extreme learning machine-based robot scene recognition method |
| CN105718963A (en) * | 2016-03-09 | 2016-06-29 | 东南大学 | SAR image classification method based on variable-length incremental type extreme learning machine |
| US9785886B1 (en) * | 2017-04-17 | 2017-10-10 | SparkCognition, Inc. | Cooperative execution of a genetic algorithm with an efficient training algorithm for data-driven model creation |
| CN107679543A (en) * | 2017-02-22 | 2018-02-09 | 天津大学 | Sparse autocoder and extreme learning machine stereo image quality evaluation method |
| CN107798383A (en) * | 2017-10-27 | 2018-03-13 | 天津大学 | Improved core extreme learning machine localization method |
-
2018
- 2018-03-30 CN CN201810295090.XA patent/CN108664992B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103400190A (en) * | 2013-08-13 | 2013-11-20 | 浙江大学 | Integrated framework method for optimizing extremity learning machine by using genetic algorithm |
| CN104537382A (en) * | 2015-01-12 | 2015-04-22 | 杭州电子科技大学 | Electromyographic signal gait recognition method for optimizing support vector machine based on genetic algorithm |
| CN104657718A (en) * | 2015-02-13 | 2015-05-27 | 武汉工程大学 | Face recognition method based on face image feature extreme learning machine |
| CN104700078A (en) * | 2015-02-13 | 2015-06-10 | 武汉工程大学 | Scale-invariant feature extreme learning machine-based robot scene recognition method |
| CN105718963A (en) * | 2016-03-09 | 2016-06-29 | 东南大学 | SAR image classification method based on variable-length incremental type extreme learning machine |
| CN107679543A (en) * | 2017-02-22 | 2018-02-09 | 天津大学 | Sparse autocoder and extreme learning machine stereo image quality evaluation method |
| US9785886B1 (en) * | 2017-04-17 | 2017-10-10 | SparkCognition, Inc. | Cooperative execution of a genetic algorithm with an efficient training algorithm for data-driven model creation |
| CN107798383A (en) * | 2017-10-27 | 2018-03-13 | 天津大学 | Improved core extreme learning machine localization method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108664992A (en) | 2018-10-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI677852B (en) | A method and apparatus, electronic equipment, computer readable storage medium for extracting image feature | |
| CN110781406B (en) | Social network user multi-attribute inference method based on variational automatic encoder | |
| CN110163647A (en) | A kind of data processing method and device | |
| CN108536784B (en) | Comment information sentiment analysis method and device, computer storage medium and server | |
| CN110991474A (en) | Machine learning modeling platform | |
| CN113220886A (en) | Text classification method, text classification model training method and related equipment | |
| CN110135681A (en) | Risk subscribers recognition methods, device, readable storage medium storing program for executing and terminal device | |
| CN112632984A (en) | Graph model mobile application classification method based on description text word frequency | |
| JP6172317B2 (en) | Method and apparatus for mixed model selection | |
| CN110472659B (en) | Data processing method, device, computer readable storage medium and computer equipment | |
| CN110990523A (en) | Legal document determining method and system | |
| CN111709225A (en) | Event cause and effect relationship judging method and device and computer readable storage medium | |
| US9053434B2 (en) | Determining an obverse weight | |
| CN112835798B (en) | Clustering learning method, testing step clustering method and related devices | |
| CN112749737A (en) | Image classification method and device, electronic equipment and storage medium | |
| CN108664992B (en) | Classification method and device based on genetic optimization and kernel extreme learning machine | |
| CN111611228A (en) | Load balance adjustment method and device based on distributed database | |
| CN110717787A (en) | User classification method and device | |
| CN115422346A (en) | Knowledge graph-based multi-turn topic recommendation method | |
| CN113627513A (en) | Training data generation method and system, electronic device and storage medium | |
| CN113221995A (en) | Data classification method, equipment and device based on semi-supervised deep classification algorithm | |
| CN112463964A (en) | Text classification and model training method, device, equipment and storage medium | |
| CN117786234B (en) | Multimode resource recommendation method based on two-stage comparison learning | |
| CN116185843B (en) | Two-stage neural network testing method and device based on neuron coverage rate guidance | |
| CN111143552B (en) | Text information category prediction method and device and server |
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 |