Educational and teaching resources dynamic allocation management system and method based on machine learning

Abstract

The present invention discloses a dynamic allocation management system and method for educational and teaching resources based on machine learning, and relates to the field of resource allocation management. The present invention performs clustering operations on a current resource feature data set matrix and a current demander feature data set matrix to obtain feature classification center data that can reflect each educational and teaching resource and each educational and teaching resource demander, thereby providing data support for subsequent matching of educational and teaching resources with educational and teaching resource demanders; by constructing a final student grade mapping equation, a mapping tool is provided for subsequent matching of educational and teaching resources with educational and teaching resource demanders; if the number of optimal student grade data that meets the requirements does not meet the requirements, the cluster center data obtained by the above clustering operation is re-optimized, so that most types of educational and teaching resource demanders can be matched with the most suitable type of educational and teaching resources.

Description

Technical Field

[0001] The present invention belongs to the field of resource allocation management, and more specifically, particularly relates to a system and method for dynamic allocation management of educational and teaching resources based on machine learning. Background Art

[0002] Chinese patent CN111861265B discloses an online teaching resource allocation method and computer system based on a global search algorithm, including obtaining class information, resource information, course information and class hours and integrating and classifying the resource information, performing dimensionality reduction processing on the initialization data; iteratively optimizing the allocation relationship and order between teaching classes, teaching courses and teaching resources to obtain allocation results; and statistically collating the allocation results to form an allocation list and then outputting it.

[0003] With the popularization of the Internet, the diversity of students' learning methods and corresponding types of educational and teaching resources has increased significantly. However, each type of educational and teaching resource is suitable for different student groups. If the most suitable educational and teaching resources cannot be allocated to different types of educational and teaching resource demanders, it will lead to poor learning effects and thus affect academic performance. Summary of the invention

[0004] In response to the problems in the related technology, the present invention proposes a dynamic allocation management system and method for educational and teaching resources based on machine learning to overcome the above-mentioned technical problems existing in the existing related technology.

[0005] In order to solve the above technical problems, the present invention is achieved through the following technical solutions:

[0006] The present invention is a method for dynamic allocation and management of educational and teaching resources based on machine learning, comprising the following steps:

[0007] S1. Collect characteristic data corresponding to the current multiple types of educational and teaching resources and types of educational and teaching resource demanders to obtain a current resource characteristic data set matrix and a current demander characteristic data set matrix;

[0008] S2, performing clustering operation on the current resource feature data set matrix and the current demander feature data set matrix to obtain a resource feature data classification center matrix set and a demander feature data classification center matrix set;

[0009] S3, collect characteristic data of various types of educational and teaching resources and types of educational and teaching resource demanders and corresponding historical student grade data in cooperation with the current resource characteristic data set matrix and the current demander characteristic data set matrix, and obtain the historical resource characteristic data matrix, the historical demander characteristic data matrix and the student grade data set; and then construct the final student grade mapping equation according to the historical resource characteristic data matrix, the historical demander characteristic data matrix and the student grade data set;

[0010] S4. Map and screen the resource feature data classification center matrix set and the demander feature data classification center matrix set according to the final student grade mapping equation to obtain a resource feature matching data matrix set and an optimal student grade mapping data matrix; optimize the resource feature data classification center matrix set and the demander feature data classification center matrix set according to the optimal student grade mapping data matrix;

[0011] Since there are many types of educational and teaching resources, the best usage scenarios of different types of educational and teaching resources are also different; correspondingly, the demanders of teaching resources have different conditions, and if they want to achieve the best teaching effect, they need different types of educational and teaching resources; based on the above, this solution defines the characteristics of educational and teaching resources and the characteristics of educational and teaching resource demanders to distinguish different types of educational and teaching resources and educational and teaching resource demanders; and collects characteristic data of different types of educational and teaching resources and educational and teaching resource demanders based on the defined educational and teaching resource types, educational and teaching resource feature types, educational and teaching demander types, and educational and teaching demander feature types, providing a data basis for subsequent clustering operations; by performing clustering operations on the current resource feature data set matrix and the current demander feature data set matrix, the characteristic classification center data that can reflect each educational and teaching resource and each educational and teaching resource demander is obtained, which not only reduces the complexity of the data, but also provides a basis for subsequent educational and teaching operations. It provides data support for matching resources with education and teaching resource demanders; since students' grades can directly reflect whether current education and teaching resources match current education and teaching resource demanders, this scheme collects characteristic data of various historical education and teaching resource types, education and teaching resource demander types and corresponding historical student grade data, and uses these data to optimize the constructed student grade mapping equation, so that the final student grade mapping equation can accurately map the characteristic data of education and teaching resource types and education and teaching resource demander types with student grade data, providing a mapping tool for subsequent matching of education and teaching resources with education and teaching resource demanders; when the matching is completed, if the number of optimal student grade data that meets the requirements does not meet the requirements, the cluster center data obtained by the above clustering operation is re-optimized, so that most types of education and teaching resource demanders can be matched with the most suitable education and teaching resource types.

[0012] Preferably, the S1 comprises the following steps:

[0013] S11. Set a plurality of educational and teaching resource types and a plurality of corresponding feature types to obtain a resource type set and a resource feature type set;

[0014] Then, multiple types of education and teaching resource demanders and corresponding multiple characteristic types are set to obtain a resource demander type set and a demander characteristic type set;

[0015] S12, collecting characteristic data of various types of current educational and teaching resources according to the resource type set and the resource characteristic type set to obtain a current resource characteristic data set matrix;

[0016] S13, collecting characteristic data of current multiple types of education and teaching resource demanders according to the resource demander type set and the demander characteristic type set, and obtaining a current demander characteristic data set matrix;

[0017] Resource types include textbooks, reference books, extracurricular readings, online courses, math problem-solving software, language learning applications, e-books, academic papers, journal articles, educational videos, audio lectures, animations, virtual laboratories, science laboratories, art studios, internship bases, professional guidance, coaching and feedback from teachers, learning management systems, gamified learning resources, study groups, online forums and academic communities, etc.; demander characteristics include students, lifelong learners, people with special education needs, etc.; resource feature types include text proportion features, picture proportion features, sound proportion features, video proportion features, interactive media proportion features, digital proportion features, physical proportion features, human proportion features and environmental proportion features, etc.; demander characteristics include systematic and comprehensive features, empirical research orientation features, combination of theory and practice features, focus on educational equity features, technology integration capabilities, data-driven decision support features, interdisciplinary perspective features, policy sensitivity features, continuous updating and iteration features and international vision features, etc.;

[0018] Preferably, S2 comprises the following steps:

[0019] S21, performing a clustering operation on the characteristic data set of each type of educational and teaching resources in the resource characteristic data set matrix to obtain a resource characteristic data classification center matrix set;

[0020] S22, performing a clustering operation on the characteristic data set of each type of educational resource demander in the demander characteristic data set matrix to obtain a demander characteristic data classification center matrix set;

[0021] There are different types of educational and teaching resources and educational and teaching resource demanders of the same type. These different types lead to different demands for educational and teaching resources even for the same type of educational and teaching resource demanders. Therefore, this scheme clusters the multiple educational and teaching resource feature data and educational and teaching resource demander feature data collected in each type of educational and teaching resource type and educational and teaching resource demander type, so as to provide a data basis for the subsequent matching between demanders and resources.

[0022] Preferably, the S21 comprises the following steps:

[0023] S211, setting a first initial cluster quantity set;

[0024] S212, performing clustering operations on the feature data sets of each type of educational and teaching resources in the resource feature data set matrix according to the first initial cluster quantity set and using the K-means algorithm to obtain a resource feature data classification center matrix set;

[0025] Preferably, the S22 comprises the following steps:

[0026] S221, setting a second initial cluster quantity set;

[0027] S222, performing clustering operations on the characteristic data sets of each type of educational resource demanders in the demander characteristic data set matrix according to the second initial cluster quantity set and using the K-means algorithm to obtain a demander characteristic data classification center matrix set;

[0028] The principle of the K-means algorithm is relatively simple, and it can still maintain high efficiency when processing large amounts of data. It can produce compact and well-separated clustering results. It mainly requires the user to specify the number of clusters, which makes the adjustment and application of the algorithm relatively simple. Based on the above advantages, this solution uses the K-means algorithm to perform clustering operations on the resource feature data set matrix and the demander feature data set matrix.

[0029] Preferably, S3 comprises the following steps:

[0030] S31, setting multiple student grade types to obtain a student grade type set;

[0031] S32, according to the student grade type set, resource type set, resource feature type set, resource demander type set and demander feature type set, collect feature data of various types of historical educational and teaching resources, feature data of corresponding educational and teaching resource demanders and corresponding student grade data, and obtain a historical resource feature data matrix, a historical demander feature data matrix and a student grade data set;

[0032] S33, constructing an initial student grade mapping equation; optimizing the initial student grade mapping equation using the historical resource feature data matrix, the historical demander feature data matrix, and the student grade data set to obtain a final student grade mapping equation;

[0033] Preferably, the S33 comprises the following steps:

[0034] S331, constructing the initial student grade mapping equation;

[0035] S332, substituting each row of data in the historical resource feature data matrix and the historical demander feature data matrix into the initial student grade mapping equation to obtain an initial student grade mapping data set;

[0036] S333, when there is student grade mapping data in the initial student grade mapping data set that is different from the corresponding student grade data in the student grade data set, optimizing the initial student grade mapping equation until there is no student grade mapping data in the initial student grade mapping data set that is different from the corresponding student grade data in the student grade data set, thereby obtaining a final student grade mapping equation;

[0037] Otherwise, there is no need to optimize the initial student grade mapping equation, and the initial student grade mapping equation is used as the final student grade mapping equation;

[0038] Since the coefficients of the independent variables and the bias of the constructed initial student grade mapping equation are pre-set, there is a possibility of inaccurate mapping; therefore, in this scheme, the accuracy of the initial student grade mapping equation is verified by inputting the existing historical resource feature data matrix, the historical demander feature data matrix and the student grade data set into the initial student grade mapping equation. When the verification result shows inaccuracy, the coefficients of the independent variables and the bias are optimized to make the mapping more accurate.

[0039] Preferably, optimizing the initial student grade mapping equation in S333 includes the following steps:

[0040] S3331, constructing a first spotted hyena population; setting the maximum number of iterations of the first spotted hyena population to The current number of iterations is They are respectively recorded as the first maximum number of iterations and the first current number of iterations;

[0041] S3332, setting the value interval of each independent variable coefficient and the bias in the initial student grade mapping equation to obtain an independent variable value interval set and a bias value interval;

[0042] The initial position of each spotted hyena in the first spotted hyena population is set according to the independent variable value interval set and the bias value interval to obtain a first initial position matrix;

[0043] S3333, constructing a fitness function of the first spotted hyena population according to the initial student grade mapping data set and the student grade data set;

[0044] S3334, start iteration, and set the first current iteration number to 1 before iteration; in the first round of iteration, use the fitness function of the first spotted hyena population to calculate the fitness value of the initial position of each spotted hyena in the first initial position matrix to obtain a first fitness value set; use the maximum fitness value in the first fitness value set and the initial position of the corresponding spotted hyena as the first global optimal fitness value and the first global optimal position respectively; update the initial position of each spotted hyena in the first initial position matrix according to the first global optimal fitness value and the first global optimal position; after the update is completed, add 1 to the first current iteration number and enter the next round of iteration;

[0045] In each other round of iteration, the fitness function of the first spotted hyena population is used to calculate the fitness value of the position of each spotted hyena updated in the previous round of iteration to obtain a second fitness value set; the maximum fitness value in the second fitness value set and the initial position of the corresponding spotted hyena are respectively used as the second global optimal fitness value and the second global optimal position; the position of each spotted hyena updated in the previous round of iteration is updated again according to the second global optimal fitness value and the second global optimal position; after the update is completed, the first current iteration number is increased by 1 and the next round of iteration is entered;

[0046] S3335, when When , stop the iteration and get the first final global optimal position, otherwise, continue to iterate until until the time; substitute the first final global optimal position into the initial student grade mapping equation to obtain the optimized student grade mapping equation; substitute each row of data in the historical resource feature data matrix and the historical demander feature data matrix into the optimized student grade mapping equation to obtain the optimized student grade mapping data set;

[0047] When there is no student grade mapping data in the initial student grade mapping data set that is different from the corresponding student grade data in the student grade data set, the optimized student grade mapping equation is used as the final student grade mapping equation; otherwise, return to S3334 to continue iterating until there is no student grade mapping data in the initial student grade mapping data set that is different from the corresponding student grade data in the student grade data set;

[0048] The spotted hyena optimization algorithm can effectively explore the solution space by simulating the hunting behavior of spotted hyenas, which helps the algorithm to jump out of the local optimal solution and find the global optimal solution; and it can maintain population diversity during the search process, which helps to avoid premature convergence to the local optimal solution during the search process; Based on the above advantages, this scheme uses the spotted hyena optimization algorithm to iterate the initial student grade mapping equation for each independent variable coefficient and bias multiple times, and uses the accuracy of its mapping as the fitness function. Therefore, as the iteration proceeds, the accuracy of its mapping becomes higher and higher, and finally meets the requirements.

[0049] Preferably, the step S4 includes the following steps:

[0050] S41, randomly combining each resource feature data classification center data set in the resource feature data classification center matrix set and each demander feature data classification center data set in the demander feature data classification center matrix set and inputting them into the final student grade mapping equation to obtain the current student grade mapping data matrix; selecting the resource feature data classification center data set that best matches each demander feature data classification center data set according to the current student grade mapping data matrix to obtain a resource feature matching data matrix set and a corresponding optimal student grade mapping data matrix;

[0051] S42, setting a student grade threshold; optimizing and adjusting the resource feature data classification center matrix set and the demander feature data classification center matrix set according to the student grade threshold;

[0052] By randomly combining each resource feature data classification center data set in the resource feature data classification center matrix set and each demander feature data classification center data set in the demander feature data classification center matrix set and inputting them into the final student grade mapping equation for mapping, the corresponding student grade data is obtained. The higher the student grade data, the higher the match between the current resource feature data classification center data set and the current demander feature data classification center matrix set, thereby providing a basis for allocating educational and teaching resources.

[0053] Preferably, the S42 comprises the following steps:

[0054] S421, setting a qualified data amount threshold; when the number of student grade mapping data greater than or equal to the student grade threshold in the optimal student grade mapping data matrix is ​​less than the qualified data amount threshold, optimizing and adjusting the resource feature data classification center matrix set and the demander feature data classification center matrix set until there is student grade mapping data greater than or equal to the student grade threshold in the optimal student grade mapping data matrix;

[0055] Otherwise, there is no need to optimize and adjust the resource feature data classification center matrix set and the demander feature data classification center matrix set;

[0056] When the total number of matched best student grade data that meets the requirements does not meet the requirements, it indicates that the multiple classification center data obtained by clustering the resource feature data set matrix and the demander feature data set matrix are not accurate enough. Therefore, the K value set by the K-means algorithm is fed back through the total number of best student grade data that meets the requirements so that the K value can be adjusted in a timely and targeted manner.

[0057] Preferably, in S421, optimizing and adjusting the resource feature data classification center matrix set and the demander feature data classification center matrix set until there is student grade mapping data greater than or equal to the student grade threshold in the optimal student grade mapping data matrix includes the following steps:

[0058] S4211, construct a second spotted hyena population; set the maximum number of iterations of the second spotted hyena population to The current number of iterations is are recorded as the second maximum number of iterations and the second current number of iterations respectively;

[0059] S4212: Set a value interval of each cluster number in the first initial cluster number set and the second initial cluster number set to obtain a first cluster value interval set and a second cluster value interval set;

[0060] The initial position of each spotted hyena in the second spotted hyena population is set according to the first cluster value interval set and the second cluster value interval set to obtain an initial position matrix set;

[0061] S4213, setting the fitness function of the second spotted hyena population according to the optimal student grade mapping data matrix and the student grade threshold;

[0062] S4214, start iteration, and set the second current iteration number to 1 before iteration; in the first round of iteration, use the fitness function of the second spotted hyena population to calculate the fitness value of the initial position matrix of each spotted hyena in the initial position matrix set to obtain a third fitness value set; use the maximum fitness value in the third fitness value set and the initial position of the corresponding spotted hyena as the third global optimal fitness value and the third global optimal position respectively; update the initial position matrix of each spotted hyena in the initial position matrix set according to the third global optimal fitness value and the third global optimal position; after the update is completed, add 1 to the second current iteration number and enter the next round of iteration;

[0063] In each other round of iteration, the fitness function of the second spotted hyena population is used to calculate the fitness value of the position matrix of each spotted hyena updated in the previous round of iteration, to obtain a fourth fitness value set; the maximum fitness value in the fourth fitness value set and the initial position of the corresponding spotted hyena are respectively used as the fourth global optimal fitness value and the fourth global optimal position; the position matrix of each spotted hyena updated in the previous round of iteration is updated again according to the fourth global optimal fitness value and the fourth global optimal position; after the update is completed, the second current iteration number is increased by 1 and the next round of iteration is entered;

[0064] S4215, when When , stop the iteration and get the second final global optimal position, otherwise, continue to iterate until until the second final global optimal position is used to replace the first initial cluster number set and the second initial cluster number set;

[0065] The spotted hyena optimization algorithm is used to iteratively optimize the number of clusters set by the K-means algorithm, and the total number of best student grade data that meets the requirements is used as the fitness function. Therefore, as the iterations proceed, the total number of best student grade data that meets the requirements will become higher and higher, and finally meet the requirements.

[0066] The dynamic allocation and management system of educational and teaching resources based on machine learning includes a current educational and teaching resource feature data collection module, a current educational and teaching resource demander feature data collection module, a clustering module, a history educational and teaching resource feature data collection module, a history educational and teaching resource demander feature data collection module, a history grade data collection module, a mapping equation construction module, an educational and teaching resource allocation screening module and a clustering optimization module.

[0067] The present invention has the following beneficial effects:

[0068] 1. In the present invention, clustering operations are performed on the current resource feature data set matrix and the current demander feature data set matrix to obtain feature classification center data that can reflect each educational and teaching resource and each educational and teaching resource demander, thereby providing data support for the subsequent matching of educational and teaching resources with educational and teaching resource demanders; by constructing the final student grade mapping equation, a mapping tool is provided for the subsequent matching of educational and teaching resources with educational and teaching resource demanders; when the matching is completed, if the number of optimal student grade data that meets the requirements does not meet the requirements, the cluster center data obtained by the above clustering operation is re-optimized, so that most types of educational and teaching resource demanders can be matched with the most suitable type of educational and teaching resources.

[0069] 2. In the present invention, the total number of best student grade data that meets the requirements is used to provide feedback to the K value set by the K-means algorithm, so as to make timely and targeted adjustments to the K value.

[0070] 3. In the present invention, the spotted hyena optimization algorithm is used to iterate the coefficients of the independent variables, the bias and the number of clusters set by the K-means algorithm of the initial student grade mapping equation for multiple times, and the mapping accuracy and the total number of the best student grade data that meet the requirements are used as fitness functions. Therefore, as the iteration proceeds, the mapping accuracy and the total number of the best student grade data that meet the requirements become higher and higher, and finally meet the requirements.

[0071] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, they can also obtain drawings based on these drawings without paying creative work.

[0073] Figure 1 It is a flow chart of the method for dynamic allocation and management of educational and teaching resources based on machine learning of the present invention. DETAILED DESCRIPTION

[0074] The following will be combined with the drawings in the embodiments of the invention to clearly and completely describe the technical solutions in the embodiments of the invention. Obviously, the described embodiments are only part of the embodiments of the invention, not all of the embodiments. Based on the embodiments in the invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the invention.

[0075] In the description of the present invention, it is necessary to understand that the terms "opening", "upper", "lower", "top", "middle", "inside" and the like indicating orientation or positional relationship are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the invention.

[0076] Embodiment 1

[0077] This embodiment is a method for dynamic allocation and management of educational and teaching resources based on machine learning, comprising the following steps:

[0078] S1. Collect characteristic data corresponding to the current multiple types of educational and teaching resources and types of educational and teaching resource demanders to obtain a current resource characteristic data set matrix and a current demander characteristic data set matrix;

[0079] The S1 comprises the following steps:

[0080] S11. Set multiple types of educational and teaching resources and corresponding multiple feature types to obtain a resource type set. and resource feature type sets a 1i represents the set i-th type of educational and teaching resources, a1′ represents the total number of set educational and teaching resource types; a 2i represents the set i-th educational and teaching resource feature type, a′2 represents the total number of set educational and teaching resource feature types;

[0081] Then set multiple types of educational and teaching resource demanders and corresponding multiple feature types to obtain the resource demander type set And the demander feature type set a 3i represents the set i-th type of educational and teaching resource demander, a3′ represents the total number of the set educational and teaching resource demander types; a 4i represents the set i-th characteristic type of educational and teaching resource demanders, a′4 represents the total number of set characteristic types of educational and teaching resource demanders;

[0082] S12: Collect feature data of various types of current educational resources according to the resource type set and resource feature type set to obtain a current resource feature data set matrix. as follows,

[0083]

[0084] in, represents the feature data set of the jth educational and teaching resource currently collected for the i-th type of educational and teaching resource, Indicates the total number of characteristic data sets of educational and teaching resources currently collected for each type of educational and teaching resources; Indicates the characteristic data of the kth type of the jth educational and teaching resource currently collected for the ith type of educational and teaching resource;

[0085] S13: Collect characteristic data of various types of education and teaching resource demanders according to the resource demander type set and demander characteristic type set to obtain a current demander characteristic data set matrix. as follows,

[0086]

[0087] in, represents the characteristic data set of the jth educational and teaching resource demander currently collected for the i-th type of educational and teaching resource demander, Indicates the total number of characteristic data sets of educational and teaching resource demanders currently collected for each type of educational and teaching resource demander; express The characteristic data of the kth type of educational and teaching resource demanders;

[0088] S2, performing clustering operation on the current resource feature data set matrix and the current demander feature data set matrix to obtain a resource feature data classification center matrix set and a demander feature data classification center matrix set;

[0089] The S2 comprises the following steps:

[0090] S21, clustering the characteristic data set of each type of educational and teaching resources in the resource characteristic data set matrix to obtain a resource characteristic data classification center matrix set b i The characteristic data classification center matrix representing the i-th type of educational and teaching resources in the resource characteristic data set matrix is ​​as follows:

[0091]

[0092] Among them, b ijk represents the kth type of feature data in the jth feature data classification center data set of the ith type of educational and teaching resources in the resource feature data set matrix, b i ′ represents the total number of the i-th type of feature data classification center data sets in the resource feature data set matrix;

[0093] The S21 comprises the following steps:

[0094] S211, setting the first initial cluster number set

[0095] S212, performing clustering operations on the feature data sets of each type of educational and teaching resources in the resource feature data set matrix according to the first initial cluster quantity set and using the K-means algorithm to obtain a resource feature data classification center matrix set;

[0096] S22, clustering the characteristic data set of each type of educational resource demander in the demander characteristic data set matrix to obtain a demander characteristic data classification center matrix set c i The characteristic data classification center matrix of the i-th type of educational resource demander in the demander characteristic data set matrix is ​​represented as follows:

[0097]

[0098] Among them, c ijk represents the kth type of characteristic data in the classification center data set of the jth characteristic data of the education and teaching resource demander of the ith type in the demander characteristic data set matrix, c i ′ represents the total number of the i-th type of feature data classification center data sets in the demander feature data set matrix;

[0099] The S22 comprises the following steps:

[0100] S221, setting the second initial cluster quantity set

[0101] S222, performing clustering operations on the characteristic data sets of each type of educational resource demanders in the demander characteristic data set matrix according to the second initial cluster quantity set and using the K-means algorithm to obtain a demander characteristic data classification center matrix set;

[0102] S3, collect characteristic data of various types of educational and teaching resources and types of educational and teaching resource demanders and corresponding historical student grade data in cooperation with the current resource characteristic data set matrix and the current demander characteristic data set matrix, and obtain the historical resource characteristic data matrix, the historical demander characteristic data matrix and the student grade data set; and then construct the final student grade mapping equation according to the historical resource characteristic data matrix, the historical demander characteristic data matrix and the student grade data set;

[0103] The S3 comprises the following steps:

[0104] S31. Set multiple student grade types to obtain a set of student grade types. Indicates the set grade type of the i-th student. Indicates the total number of student grade level types that are set;

[0105] S32, according to the student grade type set, resource type set, resource feature type set, resource demander type set and demander feature type set, collect the feature data of various types of historical educational and teaching resources, the feature data of the corresponding educational and teaching resource demanders and the corresponding student grade data, and obtain the historical resource feature data matrix d1, the historical demander feature data matrix d2 and the student grade data set d3 = {d 31 ,d 32 ,...,d 3i ,...,d 3d′}, d 3irepresents the i-th historical student grade data collected, d′ represents the total number of historical student grade data collected; d1 and d2 are as follows:

[0106]

[0107] Among them, d 1ij represents the jth type of characteristic data of the educational and teaching resources corresponding to the i-th historical student grade data collected, d 2ij Indicates the j-th type of characteristic data of the educational and teaching resource demander corresponding to the i-th historical student grade data collected;

[0108] S33, constructing an initial student grade mapping equation; optimizing the initial student grade mapping equation using the historical resource feature data matrix d1, the historical demander feature data matrix d2, and the student grade data set to obtain a final student grade mapping equation;

[0109] The S33 comprises the following steps:

[0110] S331, construct the initial student grade mapping equation; as follows,

[0111]

[0112] In the formula, is the dependent variable of the initial student grade mapping equation, representing the student grade data; is the jth independent variable of the initial student grade mapping equation, representing the jth type of characteristic data of educational and teaching resources; for The corresponding coefficients of the independent variables; is the (a′2+j)th independent variable of the initial student grade mapping equation, representing the jth type of characteristic data of the educational and teaching resource demander; for The corresponding independent variable coefficient; d5 represents the bias of the initial student grade mapping equation;

[0113] S332: Substitute each row of data in the historical resource feature data matrix and the historical demander feature data matrix into the initial student grade mapping equation to obtain an initial student grade mapping data set.

[0114] e={e1,e2,...,e i ,...,e d′},e iIndicates the i-th student grade mapping data obtained by substituting each row of data in the historical resource feature data matrix and the historical demander feature data matrix into the initial student grade mapping equation;

[0115] S333, when there is student grade mapping data in the initial student grade mapping data set that is different from the corresponding student grade data in the student grade data set, optimizing the initial student grade mapping equation until there is no student grade mapping data in the initial student grade mapping data set that is different from the corresponding student grade data in the student grade data set, thereby obtaining a final student grade mapping equation;

[0116] Otherwise, there is no need to optimize the initial student grade mapping equation, and the initial student grade mapping equation is used as the final student grade mapping equation;

[0117] Optimizing the initial student grade mapping equation in S333 includes the following steps:

[0118] S3331. Establishing the first spotted hyena population e1′ i represents the i-th spotted hyena in the first spotted hyena population, represents the size of the first spotted hyena population; sets the maximum number of iterations of the first spotted hyena population to The current number of iterations is are respectively recorded as the first maximum number of iterations and the first current number of iterations; the search space dimension of the first spotted hyena population is a′2+a′4+1;

[0119] S3332, set the value interval of each independent variable coefficient and the bias in the initial student grade mapping equation to obtain the independent variable value interval set f1 and the bias value interval They represent the lower limit and upper limit of the bias value in the initial student grade mapping equation respectively; f1 is as follows,

[0120]

[0121] in, Respectively represent the lower limit and upper limit of the value of the ith independent variable in the initial student grade mapping equation;

[0122] The initial position of each spotted hyena in the first spotted hyena population is set according to the independent variable value interval set f1 and the offset value interval to obtain a first initial position matrix f′; as follows,

[0123]

[0124] Among them, f ij ′ represents the component of the initial position of the i-th spotted hyena in the first spotted hyena population in the j-th search space dimension; f ij '、 The calculation formulas are as follows:

[0125]

[0126] In the formula, rand ij , Respectively for f ij '、 Generate a random number between 0 and 1;

[0127] S3333: constructing a fitness function of the first spotted hyena population according to the initial student grade mapping data set and the student grade data set as follows,

[0128]

[0129] Among them, β is a positive number, indicating the correction parameter;

[0130] S3334, start iteration, before iteration, set the first current iteration number Set to 1; the fitness function of the first spotted hyena population is used in the first iteration Calculate the fitness value of the initial position of each spotted hyena in the first initial position matrix to obtain a first fitness value set; use the maximum fitness value in the first fitness value set and the initial position of the corresponding spotted hyena as the first global optimal fitness value and the first global optimal position respectively; update the initial position of each spotted hyena in the first initial position matrix according to the first global optimal fitness value and the first global optimal position; after the update is completed, the first current iteration number Add 1 and enter the next iteration;

[0131] In each other iteration, the fitness function of the first spotted hyena population is used Calculate the fitness value of the position of each spotted hyena updated in the previous round of iterations to obtain a second fitness value set; use the maximum fitness value in the second fitness value set and the initial position of the corresponding spotted hyena as the second global optimal fitness value and the second global optimal position respectively; update the position of each spotted hyena updated in the previous round of iterations according to the second global optimal fitness value and the second global optimal position; after the update is completed, the first current iteration number Add 1 and enter the next iteration;

[0132] S3335, when When , stop the iteration and get the first final global optimal position, otherwise, continue to iterate until until the time; substitute the first final global optimal position into the initial student grade mapping equation to obtain the optimized student grade mapping equation; substitute each row of data in the historical resource feature data matrix and the historical demander feature data matrix into the optimized student grade mapping equation to obtain the optimized student grade mapping data set;

[0133] When there is no student grade mapping data in the initial student grade mapping data set that is different from the corresponding student grade data in the student grade data set, the optimized student grade mapping equation is used as the final student grade mapping equation; otherwise, return to S3334 to continue iterating until there is no student grade mapping data in the initial student grade mapping data set that is different from the corresponding student grade data in the student grade data set;

[0134] S4. Map and screen the resource feature data classification center matrix set and the demander feature data classification center matrix set according to the final student grade mapping equation to obtain a resource feature matching data matrix set and an optimal student grade mapping data matrix; optimize the resource feature data classification center matrix set and the demander feature data classification center matrix set according to the optimal student grade mapping data matrix;

[0135] The shown S4 comprises the following steps:

[0136] S41, randomly combine each resource feature data classification center data set in the resource feature data classification center matrix set and each demander feature data classification center data set in the demander feature data classification center matrix set and input them into the final student grade mapping equation to obtain the current student grade mapping data matrix; select the resource feature data classification center data set that best matches each demander feature data classification center data set according to the current student grade mapping data matrix to obtain the resource feature matching data matrix set. And the corresponding optimal student grade mapping data matrix Indicates that i The most matching resource feature data classification center data matrix; They are as follows:

[0137]

[0138] in, Indicates that iThe k-th type of feature data in the resource feature data classification center data set that best matches the j-th feature data classification center data set; express The j-th feature data classification center data set and the corresponding c i The student grade mapping data obtained by mapping after matching the j-th feature data classification center data set;

[0139] S42, setting a student grade threshold; optimizing and adjusting the resource feature data classification center matrix set and the demander feature data classification center matrix set according to the student grade threshold;

[0140] The S42 comprises the following steps:

[0141] S421, setting a qualified data amount threshold; when the number of student grade mapping data greater than or equal to the student grade threshold in the optimal student grade mapping data matrix is ​​less than the qualified data amount threshold, optimizing and adjusting the resource feature data classification center matrix set and the demander feature data classification center matrix set until there is student grade mapping data greater than or equal to the student grade threshold in the optimal student grade mapping data matrix;

[0142] Otherwise, there is no need to optimize and adjust the resource feature data classification center matrix set and the demander feature data classification center matrix set;

[0143] In S421, the resource feature data classification center matrix set and the demander feature data classification center matrix set are optimized and adjusted until there is student grade mapping data greater than or equal to the student grade threshold in the optimal student grade mapping data matrix, which includes the following steps:

[0144] S4211. Establishment of the second spotted hyena population e′ 2i represents the i-th spotted hyena in the second spotted hyena population, represents the size of the second spotted hyena population; sets the maximum number of iterations of the second spotted hyena population to The current number of iterations is are respectively recorded as the second maximum number of iterations and the second current number of iterations; the search space dimension of the second spotted hyena population is a1′+a3′;

[0145] S4212, setting the value interval of each cluster number in the first initial cluster number set and the second initial cluster number set to obtain a first cluster value interval set g1 and a second cluster value interval set g2; respectively as follows,

[0146]

[0147] in, respectively represent the lower limit and the upper limit of the number of the i-th initial cluster in the first initial cluster number set; respectively represent the lower limit and the upper limit of the number of the i-th initial cluster in the second initial cluster number set;

[0148] The initial position of each spotted hyena in the second spotted hyena population is set according to the first cluster value interval set and the second cluster value interval set to obtain an initial position matrix set g′ j represents the initial position matrix of the j-th spotted hyena in the second spotted hyena population; as follows,

[0149]

[0150] Among them, g′ j1i , g′ j2i They respectively represent the components of the initial position of the j-th spotted hyena in the second spotted hyena population on the i-th initial cluster number dimension in the first initial cluster number set and the second initial cluster number set; the calculation formulas are as follows,

[0151]

[0152] In the formula, rand j1i 、rand j2i Respectively represent the j1i and g′ j2i Generate a random number between 0 and 1;

[0153] S4213, setting the fitness function of the second spotted hyena population according to the optimal student grade mapping data matrix and the student grade threshold as follows,

[0154]

[0155] In the formula, Indicates the total number of student grade mapping data in the optimal student grade mapping data matrix that is greater than or equal to the student grade threshold;

[0156] S4214, start iteration, before iteration, set the second current iteration number Set to 1; the fitness function of the second spotted hyena population is used in the first iteration Calculate the fitness value of the initial position matrix of each spotted hyena in the initial position matrix set to obtain a third fitness value set; use the maximum fitness value in the third fitness value set and the initial position of the corresponding spotted hyena as the third global optimal fitness value and the third global optimal position respectively; update the initial position matrix of each spotted hyena in the initial position matrix set according to the third global optimal fitness value and the third global optimal position; after the update is completed, set the second current iteration number Add 1 and enter the next iteration;

[0157] In each other iteration, the fitness function of the second spotted hyena population is used Calculate the fitness value of the position matrix of each spotted hyena updated in the previous round of iteration to obtain a fourth fitness value set; use the maximum fitness value in the fourth fitness value set and the initial position of the corresponding spotted hyena as the fourth global optimal fitness value and the fourth global optimal position respectively; update the position matrix of each spotted hyena updated in the previous round of iteration again according to the fourth global optimal fitness value and the fourth global optimal position; after the update is completed, the second current iteration number Add 1 and enter the next iteration;

[0158] S4215, when When , stop the iteration and get the second final global optimal position, otherwise, continue to iterate until until the time; using the second final global optimal position to replace the first initial cluster number set and the second initial cluster number set.

[0159] Embodiment 2

[0160] This embodiment discloses a dynamic allocation and management system for educational and teaching resources based on machine learning. The system can implement the method of the above embodiment, including a current educational and teaching resource feature data collection module, a current educational and teaching resource demander feature data collection module, a clustering module, a history educational and teaching resource feature data collection module, a history educational and teaching resource demander feature data collection module, a history grade data collection module, a mapping equation construction module, an educational and teaching resource allocation screening module, and a clustering optimization module;

[0161] The current educational and teaching resource feature data collection module is used to collect feature data corresponding to the current multiple educational and teaching resource types to obtain a current resource feature data set matrix;

[0162] The current education and teaching resource demander feature data collection module is used to collect feature data corresponding to the current multiple types of education and teaching resource demanders to obtain a current demander feature data set matrix;

[0163] The clustering module is used to perform clustering operations on the current resource feature data set matrix and the current demander feature data set matrix to obtain a resource feature data classification center matrix set and a demander feature data classification center matrix set;

[0164] The historical education and teaching resource feature data collection module is used to collect feature data corresponding to various types of historical education and teaching resources in cooperation with the current resource feature data set matrix to obtain a historical resource feature data matrix;

[0165] The historical education and teaching resource demander characteristic data collection module is used to collect characteristic data corresponding to various types of historical education and teaching resource demanders in cooperation with the current demander characteristic data set matrix to obtain a historical demander characteristic data matrix;

[0166] The historical grade data collection module is used to collect historical student grade data corresponding to the historical resource feature data matrix and the historical demander feature data matrix to obtain a student grade data set;

[0167] The mapping equation construction module is used to construct the final student grade mapping equation based on the historical resource feature data matrix, the historical demander feature data matrix and the student grade data set;

[0168] The educational resource allocation and screening module is used to map and screen the resource feature data classification center matrix set and the demander feature data classification center matrix set according to the final student grade mapping equation to obtain a resource feature matching data matrix set and an optimal student grade mapping data matrix;

[0169] The cluster optimization module is used to optimize the resource feature data classification center matrix set and the demander feature data classification center matrix set according to the optimal student grade mapping data matrix.

[0170] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0171] The preferred embodiments of the invention disclosed above are only used to help explain the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the invention, so that those skilled in the art can understand and use the invention well.

Claims

1. A method for dynamic allocation and management of educational and teaching resources based on machine learning, characterized in that: The following steps are involved: S1. Collect characteristic data corresponding to the current multiple types of educational and teaching resources and types of educational and teaching resource demanders to obtain a current resource characteristic data set matrix and a current demander characteristic data set matrix; S2, performing clustering operation on the current resource feature data set matrix and the current demander feature data set matrix to obtain a resource feature data classification center matrix set and a demander feature data classification center matrix set; S3, collect characteristic data of various types of educational and teaching resources and types of educational and teaching resource demanders and corresponding historical student grade data in cooperation with the current resource characteristic data set matrix and the current demander characteristic data set matrix, and obtain the historical resource characteristic data matrix, the historical demander characteristic data matrix and the student grade data set; and then construct the final student grade mapping equation according to the historical resource characteristic data matrix, the historical demander characteristic data matrix and the student grade data set; S4. Map and screen the resource feature data classification center matrix set and the demander feature data classification center matrix set according to the final student grade mapping equation to obtain a resource feature matching data matrix set and an optimal student grade mapping data matrix; optimize the resource feature data classification center matrix set and the demander feature data classification center matrix set according to the optimal student grade mapping data matrix; The S3 comprises the following steps: S31, setting multiple student grade types to obtain a student grade type set; S32, according to the student grade type set, resource type set, resource feature type set, resource demander type set and demander feature type set, collect feature data of various types of historical educational and teaching resources, feature data of corresponding educational and teaching resource demanders and corresponding student grade data, and obtain a historical resource feature data matrix, a historical demander feature data matrix and a student grade data set; S33, constructing an initial student grade mapping equation; optimizing the initial student grade mapping equation using the historical resource feature data matrix, the historical demander feature data matrix, and the student grade data set to obtain a final student grade mapping equation; The S33 comprises the following steps: S331, constructing the initial student grade mapping equation; S332, substituting each row of data in the historical resource feature data matrix and the historical demander feature data matrix into the initial student grade mapping equation to obtain an initial student grade mapping data set; S333, when there is student grade mapping data in the initial student grade mapping data set that is different from the corresponding student grade data in the student grade data set, optimizing the initial student grade mapping equation until there is no student grade mapping data in the initial student grade mapping data set that is different from the corresponding student grade data in the student grade data set, thereby obtaining a final student grade mapping equation; Otherwise, there is no need to optimize the initial student grade mapping equation, and the initial student grade mapping equation is used as the final student grade mapping equation.

2. The method for dynamic allocation and management of educational and teaching resources based on machine learning according to claim 1 is characterized in that: The S1 comprises the following steps: S11. Set a plurality of educational and teaching resource types and a plurality of corresponding feature types to obtain a resource type set and a resource feature type set; then set a plurality of educational and teaching resource demander types and a plurality of corresponding feature types to obtain a resource demander type set and a demander feature type set; S12, collecting characteristic data of various types of current educational and teaching resources according to the resource type set and the resource characteristic type set to obtain a current resource characteristic data set matrix; S13. Collect characteristic data of various types of current education and teaching resource demanders according to the resource demander type set and the demander characteristic type set to obtain a current demander characteristic data set matrix.

3. The method for dynamic allocation and management of educational and teaching resources based on machine learning according to claim 2 is characterized in that: The S2 comprises the following steps: S21, performing a clustering operation on the characteristic data set of each type of educational and teaching resources in the resource characteristic data set matrix to obtain a resource characteristic data classification center matrix set; S22. Perform a clustering operation on the characteristic data set of each type of education and teaching resource demander in the demander characteristic data set matrix to obtain a demander characteristic data classification center matrix set.

4. The method for dynamic allocation and management of educational and teaching resources based on machine learning according to claim 3 is characterized in that: Optimizing the initial student grade mapping equation in S333 includes the following steps: S3331, constructing a first spotted hyena population; setting the maximum number of iterations of the first spotted hyena population to , the current number of iterations is , respectively recorded as the first maximum number of iterations and the first current number of iterations; S3332, setting the value interval of each independent variable coefficient and the bias in the initial student grade mapping equation to obtain an independent variable value interval set and a bias value interval; setting the initial position of each spotted hyena in the first spotted hyena population according to the independent variable value interval set and the bias value interval to obtain a first initial position matrix; S3333, constructing a fitness function of the first spotted hyena population according to the initial student grade mapping data set and the student grade data set; S3334, start iteration; in each round of iteration, use the fitness function of the first spotted hyena population to calculate the fitness value of the position of each spotted hyena updated in the previous round of iteration, and update the position of each spotted hyena updated in the previous round of iteration; S3335, when When , stop the iteration and get the first final global optimal position, otherwise, continue to iterate until until the time; substitute the first final global optimal position into the initial student grade mapping equation to obtain the optimized student grade mapping equation; substitute each row of data in the historical resource feature data matrix and the historical demander feature data matrix into the optimized student grade mapping equation to obtain the optimized student grade mapping data set; When there is no student grade mapping data in the initial student grade mapping data set that is different from the corresponding student grade data in the student grade data set, the optimized student grade mapping equation is used as the final student grade mapping equation; otherwise, return to S3334 to continue iterating until there is no student grade mapping data in the initial student grade mapping data set that is different from the corresponding student grade data in the student grade data set.

5. The method for dynamic allocation and management of educational and teaching resources based on machine learning according to claim 4 is characterized in that: The shown S4 comprises the following steps: S41, randomly combining each resource feature data classification center data set in the resource feature data classification center matrix set and each demander feature data classification center data set in the demander feature data classification center matrix set and inputting them into the final student grade mapping equation to obtain the current student grade mapping data matrix; selecting the resource feature data classification center data set that best matches each demander feature data classification center data set according to the current student grade mapping data matrix to obtain a resource feature matching data matrix set and a corresponding optimal student grade mapping data matrix; S42, setting a student grade threshold; optimizing and adjusting the resource feature data classification center matrix set and the demander feature data classification center matrix set according to the student grade threshold.

6. The method for dynamic allocation and management of educational and teaching resources based on machine learning according to claim 5 is characterized in that: The S42 comprises the following steps: S421. Set a qualified data quantity threshold; when the number of student grade mapping data greater than or equal to the student grade threshold in the optimal student grade mapping data matrix is ​​less than the qualified data quantity threshold, optimize and adjust the resource feature data classification center matrix set and the demander feature data classification center matrix set until there is student grade mapping data greater than or equal to the student grade threshold in the optimal student grade mapping data matrix; otherwise, there is no need to optimize and adjust the resource feature data classification center matrix set and the demander feature data classification center matrix set.

7. The method for dynamic allocation and management of educational and teaching resources based on machine learning according to claim 6 is characterized in that: In S421, the resource feature data classification center matrix set and the demander feature data classification center matrix set are optimized and adjusted using the spotted hyena optimization algorithm.

8. A system for implementing the method for dynamic allocation and management of educational and teaching resources based on machine learning as described in any one of claims 1 to 7.

Images