Decision generation system and method based on big data and applied to toning template manufacturing

Abstract

The invention discloses a decision generation system and method based on big data and applied to toning template manufacturing, and relates to the field of toning template manufacturing decision generation. According to feature data of a toning template, feature data of a user, feature data of a scene where the user uses toning and the degree of satisfaction of the user, a mapping model is constructed; the feature data of the current user and the feature data of the scene can be subsequently obtained; by collecting a plurality of toning template picture examples, data support is provided for subsequently constructing a standard toning template; the samples are clustered from the aspect of sample features, so that sample feature data of similar style types are clustered into one class; a plurality of alternative decisions are provided for subsequent toning decisions of a current user, a plurality of groups of historical standard toning template feature values, user feature data, scene feature data and corresponding user satisfaction level data are collected, and data support is provided for subsequent construction of a user satisfaction level mapping model.

Description

Technical Field

[0001] The present invention belongs to the field of decision-making generation for color palette template production. Specifically, it particularly relates to a decision-making generation system and method for color palette template production based on big data. Background Art

[0002] Making a color palette template can simplify the operation process. Through preset parameters and effects, it can be quickly applied to materials, greatly shortening the color adjustment time, improving work efficiency, and ensuring style consistency. However, if the color combination of the color palette template does not meet the user's aesthetic and needs, it may greatly reduce the user's first impression of the product or content, and may also give the user a negative impression of the brand, thereby affecting the long-term development of the brand; it cannot stimulate the user's interest and participation, may lead to a decrease in the user's activity in the product, and thus affect the overall operation effect of the product. Summary of the Invention

[0003] In view of the problems in the related art, the present invention proposes a decision-making generation system and method for color palette template production based on big data to overcome the above-mentioned technical problems existing in the related art.

[0004] To solve the above technical problems, the present invention is realized through the following technical solutions:

[0005] The present invention is a decision-making generation method for color palette template production based on big data, including the following steps:

[0006] S1. Collect a number of color palette template picture samples and their corresponding style types, and collect the feature data of each color palette template picture sample to obtain a color palette template picture sample feature data matrix;

[0007] S2. Classify the color palette template picture sample feature value matrix and calculate the classification center data of each classification matrix to obtain a sample feature value classification center data matrix;

[0008] S3. Collect a number of groups of historical standard color palette template feature values, user feature data, scene feature data, and their corresponding user satisfaction level data;

[0009] S4. Construct a user satisfaction level mapping model using the historical standard color palette template feature values, user feature data, scene feature data, and their corresponding user satisfaction level data;

[0010] S5. Map the current user satisfaction level data corresponding to the sample feature value classification center data matrix using the user satisfaction level mapping model to obtain a current user satisfaction level data set;

[0011] S6. Generate a current color - matching strategy based on the current user satisfaction level dataset;

[0012] In this solution, a mapping model is constructed based on the characteristic data of the color - matching template, the characteristic data of the user, the characteristic data of the scenario where the user uses color - matching, and the user's satisfaction level, so that subsequent operations can be based on the characteristic data of the current user and the characteristic data of the scenario.

[0013] Preferably, the step S1 includes the following steps:

[0014] S11. Collect a number of color - matching template picture samples and their corresponding style types on the network platform to obtain a color - matching template picture sample set and a sample style type set;

[0015] S12. Set a number of color - matching template feature types to obtain a color - matching template feature type set; collect the characteristic data of each color - matching template picture sample in the color - matching template picture sample set according to the color - matching template feature type set to obtain a color - matching template picture sample characteristic data matrix a;

[0016]

[0017] where a ij represents the j - th type of characteristic of the i - th color - matching template picture sample in the color - matching template picture sample set, and a1′ and a′2 respectively represent the total number of set color - matching template feature types and the total number of collected color - matching template picture samples;

[0018] By collecting a number of color - matching template picture samples, it provides data support for constructing a standard color - matching template in the follow - up; clustering the samples from the perspective of sample characteristics, so that the sample characteristic data of similar style types are clustered into one category; and further providing several alternative decisions for the subsequent color - matching decision of the current user.

[0019] Preferably, the step S2 includes the following steps:

[0020] S21. Numerically encode the color - matching template picture sample characteristic data matrix to obtain a color - matching template picture sample characteristic numerical matrix; classify the color - matching template picture sample characteristic numerical matrix according to the sample style type set to obtain a color - matching template picture sample characteristic numerical classification matrix set represents the i - th color - matching template picture sample characteristic numerical classification matrix obtained by classifying the color - matching template picture sample characteristic numerical matrix, represents the total number of classification matrices obtained by classifying the color - matching template picture sample characteristic numerical matrix, and the total number of set sample style types is the same as the same; As follows,

[0021]

[0022] Among them, represents the characteristic value of the k-th type of the j-th color palette template picture example in i represents being classified into the total number of color palette template picture examples in

[0023] S22. Calculate the classification center data of each color palette template picture example feature value classification matrix in the color palette template picture example feature value classification matrix set, and obtain the example feature value classification center data matrix c′; as follows,

[0024]

[0025] Among them, c i ′ k represents the classification center data of the corresponding k-th type;

[0026] By calculating the example feature value classification center data matrix, it provides data support for subsequent mapping of the feature data and scenario data of the current user, and further generates a color palette decision.

[0027] Preferably, S22 includes the following steps:

[0028] S221. Calculate the Euclidean distance between each row of data in each color palette template picture example feature value classification matrix in the color palette template picture example feature value classification matrix set and the corresponding other rows of data, and obtain the example feature value Euclidean distance matrix set represents the corresponding example feature value Euclidean distance matrix; as follows,

[0029]

[0030] Among them, represents the Euclidean distance between the j-th row of data and the k-th row of data in

[0031] S222. Take the row of data with the largest sum of data in each example feature value Euclidean distance matrix in the example feature value Euclidean distance matrix set as the classification center data of the corresponding color palette template picture example feature value classification matrix, and obtain the example feature value classification center data matrix.

[0032] Preferably, S3 includes the following steps:

[0033] S31. Take the color palette template picture examples corresponding to each row of data in the sample feature value classification center data matrix as the color palette templates for the corresponding sample styles to obtain a standard color palette template set;

[0034] S32. Set several user feature types and scenario feature types to obtain a user feature type set and a scenario feature type set; then set several user satisfaction levels to obtain a user satisfaction level set;

[0035] S33. In combination with the standard color palette template set, the sample feature value classification center data matrix, the user feature type set, the scenario feature type set, and the user satisfaction level set, collect several groups of historical standard color palette template feature values, user feature data, scenario feature data, and the corresponding user satisfaction level data to obtain a historical color palette template feature data matrix Historical user feature data matrix Historical scenario feature data matrix And a historical user satisfaction level data set Denote the i-th group of historical user satisfaction level data collected, and d denote the total number of groups of historical user satisfaction level data collected; They are respectively as follows,

[0036]

[0037] Among them, Respectively denote the j-th type of color palette template feature data, user feature data, and scenario feature data of the i-th group collected, and a3' and a'4 respectively denote the total number of user feature types and scenario feature types set;

[0038] By collecting several groups of historical standard color palette template feature values, user feature data, scenario feature data, and the corresponding user satisfaction level data, it provides data support for the subsequent construction of the user satisfaction level mapping model.

[0039] Preferably, the S4 includes the following steps:

[0040] S41. Construct an initial multi-layer perceptron model;

[0041] S42. Use the historical color palette template feature data matrix, the historical user feature data matrix, the historical scenario feature data matrix, and the historical user satisfaction level data set to train and test the initial multi-layer perceptron model to obtain a user satisfaction level mapping model;

[0042] The multi-layer perceptron model can process high-dimensional data by increasing the number of hidden layers and neurons, and has advantages in dealing with large-scale data and complex tasks; through the combination of multiple hidden layers, it can learn more generalized feature representations, has good generalization ability when dealing with unseen data, and can accurately predict new data; good fault tolerance: even if some neurons or connections fail, the multi-layer perceptron can still work normally and has a certain fault tolerance; associative memory function: it can associate and remember the input data through learning and training, so as to realize the classification and prediction of data.

[0043] Preferably, the S42 includes the following steps:

[0044] S421. Set the training data ratio; divide the historical color palette feature data matrix, historical user feature data matrix, historical scene feature data matrix, and historical user satisfaction level data set according to the training data ratio to obtain a historical color palette feature training data matrix, historical user feature training data matrix, historical scene feature training data matrix, historical user satisfaction level training data set, historical color palette feature test data matrix, historical user feature test data matrix, historical scene feature test data matrix, and historical user satisfaction level test data set;

[0045] S422. Set the training error threshold; input the historical color palette feature training data matrix, historical user feature training data matrix, and historical scene feature training data matrix as training data and the historical user satisfaction level training data set as training labels into the initial multi-layer perceptron model for training; during the training process, when the training error is less than the training error threshold, stop training to obtain a trained multi-layer perceptron model; otherwise, continue training until the training error is less than the training error threshold.

[0046] S423. Set the test accuracy threshold; input the historical color palette feature test data matrix, historical user feature test data matrix, and historical scene feature test data matrix as test data and the historical user satisfaction level test data set as test labels into the trained multi-layer perceptron model for testing to obtain test accuracy data; when the test accuracy data is greater than or equal to the test accuracy threshold, use the trained multi-layer perceptron model as the user satisfaction level mapping model; otherwise, return to S422 to continue training the trained multi-layer perceptron model until the test accuracy data is greater than or equal to the test accuracy threshold.

[0047] Preferably, the S5 includes the following steps:

[0048] S51. Collect the feature data of the current user and the scene feature data to be used for color matching in combination with the user feature type set and the scene feature type set, so as to obtain the current user feature data set and the current scene feature data set;

[0049] S52. Combine each row of data in the sample feature value classification center data matrix with the current user feature data set and the current scene feature data set and input them into the user satisfaction level mapping model for mapping to obtain the current user satisfaction level data set;

[0050] Through mapping, it provides a basis for generating color matching decisions subsequently.

[0051] Preferably, the S6 includes the following steps:

[0052] S61. Use the row data in the sample feature value classification center data matrix corresponding to the current user satisfaction level data with the largest value in the current user satisfaction level data set as the current color matching strategy.

[0053] A decision generation system for color matching template production based on big data includes an existing color matching template sample collection module, a sample feature data collection module, a sample feature data classification center calculation module, a historical user color matching data collection module, a user satisfaction level mapping model construction module, a current user satisfaction level mapping module, and a color matching decision generation module;

[0054] The existing color matching template sample collection module is used to collect several color matching template picture samples and the corresponding style types to obtain a color matching template picture sample set and a sample style type set;

[0055] The sample feature data collection module is used to collect the feature data of each color matching template picture sample in the color matching template picture sample set to obtain a color matching template picture sample feature data matrix;

[0056] The sample feature data classification center calculation module is used to classify the color matching template picture sample feature value matrix according to the sample style type set and calculate the classification center data of each classification matrix to obtain a sample feature value classification center data matrix;

[0057] The historical user color matching data collection module is used to collect several groups of historical standard color matching template feature values, user feature data, scene feature data, and the corresponding user satisfaction level data to obtain a historical color matching template feature data matrix, a historical user feature data matrix, a historical scene feature data matrix, and a historical user satisfaction level data set;

[0058] The user satisfaction level mapping model construction module is used to construct a user satisfaction level mapping model by using the historical color matching template feature data matrix, the historical user feature data matrix, the historical scenario feature data matrix, and the historical user satisfaction level data set;

[0059] The current user satisfaction level mapping module is used to map the current user satisfaction level data corresponding to the sample feature value classification center data matrix by using the user satisfaction level mapping model to obtain the current user satisfaction level data set;

[0060] The color matching decision generation module is used to generate the current color matching strategy according to the current user satisfaction level data set.

[0061] The present invention has the following beneficial effects:

[0062] 1. In the present invention, by constructing a mapping model based on the feature data of the color matching template, the feature data of the user, and the feature data of the scenario where the user uses color matching and the user's satisfaction level, subsequent processing can be carried out according to the feature data of the current user and the feature data of the scenario; thus, the finally generated color matching template production decision is more in line with the user's needs.

[0063] 2. In the present invention, by collecting a number of color matching template picture samples, it provides data support for constructing a standard color matching template subsequently; clustering the samples from the perspective of sample features, so that the sample feature data of similar style types are grouped into one category; and further providing several alternative decisions for the subsequent color matching decision of the current user.

[0064] 3. In the present invention, by collecting several groups of historical standard color matching template feature values, user feature data, scenario feature data, and corresponding user satisfaction level data, it provides data support for constructing the user satisfaction level mapping model subsequently.

[0065] Of course, it is not necessary for any product implementing the present invention to achieve all the above-mentioned advantages simultaneously. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] 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 in the following description are only some embodiments of the invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings without creative efforts.

[0067] Figure 1 It is a schematic flowchart of the decision generation method for color matching template production based on big data of the present invention;

[0068] Figure 2This is a schematic diagram of the modules of the decision-making generation system for color matching template production based on big data according to the present invention. Specific embodiments

[0069] Next, the technical solutions in the embodiments of the invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the invention. Obviously, the described embodiments are only a part of the embodiments of the invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the invention without creative efforts shall fall within the protection scope of the invention.

[0070] Embodiment 1

[0071] Please refer to Figure 1 , this embodiment is a decision-making generation method for color matching template production based on big data, including the following steps:

[0072] S1. Collect a number of color matching template picture samples and their corresponding style types, and collect the feature data of each color matching template picture sample to obtain a feature data matrix of color matching template picture samples;

[0073] The S1 includes the following steps:

[0074] S11. Collect a number of color matching template picture samples and their corresponding style types on the network platform to obtain a set of color matching template picture samples and a set of sample style types;

[0075] S12. Set a number of color matching template feature types to obtain a set of color matching template feature types; the set of color matching template feature types includes color styles, such as sugar water tone, Japanese tone, cyberpunk tone, nostalgic tone, Japanese fresh style, retro nostalgic style, Hong Kong style, and cool breeze style, etc.; data visualization types, such as qualitative color palette, sequential color palette, and diverging color palette, etc.; according to the set of color matching template feature types, collect the feature data of each color matching template picture sample in the set of color matching template picture samples to obtain a feature data matrix a of color matching template picture samples;

[0076]

[0077] Among them, a ij represents the feature of the jth type of the ith color matching template picture sample in the set of color matching template picture samples, and a1' and a'2 respectively represent the total number of set color matching template feature types and the total number of collected color matching template picture samples;

[0078] S2. Classify the feature value matrix of color matching template picture samples and calculate the classification center data of each classification matrix to obtain a classification center data matrix of sample feature values;

[0079] The said S2 includes the following steps:

[0080] S21. Numerically encode the feature data matrix of the color palette template picture sample to obtain a feature value matrix of the color palette template picture sample; classify the feature value matrix of the color palette template picture sample according to the sample style type set to obtain a set of feature value classification matrices of the color palette template picture sample represents the i-th feature value classification matrix of the color palette template picture sample obtained by classifying the feature value matrix of the color palette template picture sample, represents the total number of classification matrices obtained by classifying the feature value matrix of the color palette template picture sample, and the total number of set sample style types is the same as the same; as follows,

[0081]

[0082] Among them, represents the k-th type of feature value of the j-th color palette template picture sample in i represents classified into the total number of color palette template picture samples in;

[0083] S22. Calculate the classification center data of each feature value classification matrix in the set of feature value classification matrices of the color palette template picture sample to obtain a sample feature value classification center data matrix c′; as follows,

[0084]

[0085] Among them, c i ′ k represents the classification center data of the corresponding k-th type;

[0086] The said S22 includes the following steps:

[0087] S221. Calculate the Euclidean distance between each row of data in each feature value classification matrix in the set of feature value classification matrices of the color palette template picture sample and the corresponding other rows of data to obtain a set of sample feature value Euclidean distance matrices represents the corresponding sample feature value Euclidean distance matrix; as follows,

[0088]

[0089] Among them, represents the Euclidean distance between the j-th row of data and the k-th row of data in;

[0090] S222. Take the row of data with the largest sum of data in each Euclidean distance matrix of sample feature values as the classification center data of the corresponding color palette template image sample feature value classification matrix, and obtain a sample feature value classification center data matrix;

[0091] S3. Collect several groups of historical standard color palette template feature values, user feature data, scene feature data, and corresponding user satisfaction level data;

[0092] S3 includes the following steps:

[0093] S31. Take the color palette template images corresponding to each row of data in the sample feature value classification center data matrix as the color palette templates of the corresponding sample styles to obtain a standard color palette template set;

[0094] S32. Set several types of user feature types and scene feature types to obtain a user feature type set and a scene feature type set; the user feature type set includes age, gender, etc.; the scene feature type set includes scene themes such as shopping, entertainment, learning, work, socializing, etc.; scene elements such as product information, service content, entertainment programs, learning materials, social objects, etc.; scene layouts such as page layout, spatial layout, functional layout, etc.; then set several types of user satisfaction levels to obtain a user satisfaction level set;

[0095] S33. In cooperation with the standard color palette template set, the sample feature value classification center data matrix, the user feature type set, the scene feature type set, and the user satisfaction level set, collect several groups of historical standard color palette template feature values, user feature data, scene feature data, and corresponding user satisfaction level data to obtain a historical color palette template feature data matrix Historical user feature data matrix Historical scene feature data matrix And a historical user satisfaction level data set Represents the i-th group of historical user satisfaction level data collected, and d represents the total number of groups of historical user satisfaction level data collected; Respectively as follows,

[0096]

[0097] Among them, Respectively represent the j-th type of color palette template feature data, user feature data, and scene feature data of the i-th group collected, and a3' and a'4 respectively represent the total number of set user feature types and scene feature types;

[0098] S4. Construct a user satisfaction level mapping model using the historical standard color palette template feature values, user feature data, scene feature data, and the corresponding user satisfaction level data;

[0099] S4 includes the following steps:

[0100] S41. Construct an initial multi-layer perceptron model;

[0101] S42. Use the historical color palette template feature data matrix, historical user feature data matrix, historical scene feature data matrix, and historical user satisfaction level data set to train and test the initial multi-layer perceptron model to obtain a user satisfaction level mapping model;

[0102] S42 includes the following steps:

[0103] S421. Set the training data ratio; divide the historical color palette template feature data matrix, historical user feature data matrix, historical scene feature data matrix, and historical user satisfaction level data set according to the training data ratio to obtain a historical color palette template feature training data matrix, historical user feature training data matrix, historical scene feature training data matrix, historical user satisfaction level training data set, historical color palette template feature test data matrix, historical user feature test data matrix, historical scene feature test data matrix, and historical user satisfaction level test data set;

[0104] S422. Set the training error threshold; input the historical color palette template feature training data matrix, historical user feature training data matrix, historical scene feature training data matrix as training data and the historical user satisfaction level training data set as training labels into the initial multi-layer perceptron model for training; during the training process, when the training error is less than the training error threshold, stop training to obtain a trained multi-layer perceptron model; otherwise, continue training until the training error is less than the training error threshold;

[0105] S423. Set the test accuracy threshold; input the historical color palette template feature test data matrix, historical user feature test data matrix, historical scene feature test data matrix as test data and the historical user satisfaction level test data set as test labels into the trained multi-layer perceptron model for testing to obtain test accuracy data; when the test accuracy data is greater than or equal to the test accuracy threshold, use the trained multi-layer perceptron model as the user satisfaction level mapping model; otherwise, return to S422 to continue training the trained multi-layer perceptron model until the test accuracy data is greater than or equal to the test accuracy threshold;

[0106] S5. Map the current user satisfaction level data corresponding to the sample feature value classification center data matrix using the user satisfaction level mapping model to obtain the current user satisfaction level data set;

[0107] S5 includes the following steps:

[0108] S51. Collect the feature data of the current user and the scene feature data to be used for color matching in combination with the user feature type set and the scene feature type set to obtain the current user feature data set and the current scene feature data set;

[0109] S52. Combine each row of data in the sample feature value classification center data matrix with the current user feature data set and the current scene feature data set and input them into the user satisfaction level mapping model for mapping to obtain the current user satisfaction level data set;

[0110] S6. Generate the current color matching strategy according to the current user satisfaction level data set;

[0111] S6 includes the following steps:

[0112] S61. Use the row data in the sample feature value classification center data matrix corresponding to the current user satisfaction level data with the largest value in the current user satisfaction level data set as the current color matching strategy.

[0113] Embodiment 2

[0114] Please refer to Figure 2 , this embodiment discloses a decision-making generation system based on big data for color matching template production. The system can implement the method of the above embodiment, including an existing color matching template sample collection module, a sample feature data collection module, a sample feature data classification center calculation module, a historical user color matching data collection module, a user satisfaction level mapping model construction module, a current user satisfaction level mapping module, and a color matching decision generation module;

[0115] The existing color matching template sample collection module is used to collect a number of color matching template picture samples and the corresponding style types to obtain the color matching template picture sample set and the sample style type set;

[0116] The sample feature data collection module is used to collect the feature data of each color matching template picture sample in the color matching template picture sample set to obtain the color matching template picture sample feature data matrix;

[0117] The sample feature data classification center calculation module is used to classify the color matching template picture sample feature value matrix according to the sample style type set and calculate the classification center data of each classification matrix to obtain the sample feature value classification center data matrix;

[0118] The historical user color matching data acquisition module is used to acquire several groups of historical standard color matching template feature values, user feature data, scene feature data, and corresponding user satisfaction level data, so as to obtain a historical color matching template feature data matrix, a historical user feature data matrix, a historical scene feature data matrix, and a historical user satisfaction level data set;

[0119] The user satisfaction level mapping model construction module is used to construct a user satisfaction level mapping model by using the historical color matching template feature data matrix, the historical user feature data matrix, the historical scene feature data matrix, and the historical user satisfaction level data set;

[0120] The current user satisfaction level mapping module is used to map the current user satisfaction level data corresponding to the sample feature value classification center data matrix by using the user satisfaction level mapping model to obtain a current user satisfaction level data set;

[0121] The color matching decision generation module is used to generate a current color matching strategy according to the current user satisfaction level data set.

[0122] In the description of this specification, the descriptions referring to terms such as "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0123] 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 embodiments described. Obviously, many modifications and variations can be made according to the content of this specification. The embodiments selected and specifically described in this specification are to better explain the principle and practical application of the invention, so that those skilled in the art in the relevant technical field can understand and utilize the invention well.

Claims

1. A decision-making generation method based on big data and applied to the production of color matching templates, characterized in that It includes the following steps: S1. Collect a number of color palette template picture samples and their corresponding style types, and collect the feature data of each color palette template picture sample to obtain a color palette template picture sample feature data matrix; S2. Classify the color palette template picture sample feature value matrix and calculate the classification center data of each classification matrix to obtain a sample feature value classification center data matrix; S3. Collect several groups of historical standard color palette template feature values, user feature data, scene feature data, and their corresponding user satisfaction level data; S4. Construct a user satisfaction level mapping model using the historical standard color palette template feature values, user feature data, scene feature data, and their corresponding user satisfaction level data; S5. Map the current user satisfaction level data corresponding to the sample feature value classification center data matrix using the user satisfaction level mapping model to obtain a current user satisfaction level data set; S6. Generate a current color adjustment strategy according to the current user satisfaction level data set.

2. The decision-making generation method based on big data for making color matching templates according to claim 1, wherein The S1 includes the following steps: S11. Collect a number of color palette template picture samples and their corresponding style types on the network platform to obtain a color palette template picture sample set and a sample style type set; S12. Set a number of color palette template feature types to obtain a color palette template feature type set; collect the feature data of each color palette template picture sample in the color palette template picture sample set according to the color palette template feature type set to obtain a color palette template picture sample feature data matrix.

3. The decision generation method based on big data for making color matching templates according to claim 2, characterized in that The S2 includes the following steps: S21. Perform numerical encoding on the color palette template picture sample feature data matrix to obtain a color palette template picture sample feature value matrix; classify the color palette template picture sample feature value matrix according to the sample style type set to obtain a color palette template picture sample feature value classification matrix set; S22. Calculate the classification center data of each color palette template picture sample feature value classification matrix in the color palette template picture sample feature value classification matrix set to obtain a sample feature value classification center data matrix.

4. The decision generation method based on big data for color matching template production according to claim 3, wherein The S22 includes the following steps: S221. Calculate the Euclidean distance between each row of data in each color palette template picture sample feature value classification matrix in the color palette template picture sample feature value classification matrix set and the corresponding other rows of data to obtain a sample feature value Euclidean distance matrix set; S222. Take the row of data with the largest sum of data in each sample feature value Euclidean distance matrix in the sample feature value Euclidean distance matrix set as the classification center data of the corresponding color palette template picture sample feature value classification matrix to obtain a sample feature value classification center data matrix.

5. The decision generation method based on big data and applied to the production of color matching templates according to claim 4, characterized in that, The S3 includes the following steps: S31. Take the color palette template picture sample corresponding to each row of data in the sample feature value classification center data matrix as the color palette template of the corresponding sample style to obtain a standard color palette template set; S32. Set several user feature types and scene feature types to obtain a user feature type set and a scene feature type set; then set several user satisfaction levels to obtain a user satisfaction level set; S33. In coordination with the standard color matching template set, the sample feature value classification center data matrix, the user feature type set, the scene feature type set, and the user satisfaction level set, collect several groups of historical standard color matching template feature values, user feature data, scene feature data, and the corresponding user satisfaction level data, to obtain the historical color matching template feature data matrix, the historical user feature data matrix, the historical scene feature data matrix, and the historical user satisfaction level data set.

6. The decision-making generation method based on big data and applied to the production of color matching templates according to claim 5, characterized in that, The S4 includes the following steps: S41. Construct an initial multi-layer perceptron model; S42. Use the historical color matching template feature data matrix, the historical user feature data matrix, the historical scene feature data matrix, and the historical user satisfaction level data set to train and test the initial multi-layer perceptron model, to obtain the user satisfaction level mapping model.

7. The decision-making generation method based on big data for making color matching templates according to claim 6, characterized in that, The S42 includes the following steps: S421. Set the training data ratio; according to the training data ratio, divide the historical color matching template feature data matrix, the historical user feature data matrix, the historical scene feature data matrix, and the historical user satisfaction level data set, to obtain the historical color matching template feature training data matrix, the historical user feature training data matrix, the historical scene feature training data matrix, the historical user satisfaction level training data set, the historical color matching template feature test data matrix, the historical user feature test data matrix, the historical scene feature test data matrix, and the historical user satisfaction level test data set; S422. Set the training error threshold; input the historical color matching template feature training data matrix, the historical user feature training data matrix, and the historical scene feature training data matrix as training data, and the historical user satisfaction level training data set as the training label into the initial multi-layer perceptron model for training; during the training process, when the training error is less than the training error threshold, stop training to obtain the trained multi-layer perceptron model; otherwise, continue training until the training error is less than the training error threshold; S423. Set the test accuracy threshold; input the historical color matching template feature test data matrix, the historical user feature test data matrix, and the historical scene feature test data matrix as test data, and the historical user satisfaction level test data set as the test label into the trained multi-layer perceptron model for testing, to obtain the test accuracy data; when the test accuracy data is greater than or equal to the test accuracy threshold, use the trained multi-layer perceptron model as the user satisfaction level mapping model; otherwise, return to S422 to continue training the trained multi-layer perceptron model until the test accuracy data is greater than or equal to the test accuracy threshold.

8. The decision generation method based on big data for color matching template production according to claim 7, characterized in that The S5 includes the following steps: S51. In coordination with the user feature type set and the scene feature type set, collect the feature data of the current user and the scene feature data to be used for color matching, to obtain the current user feature data set and the current scene feature data set; S52. Combine each row of data in the sample feature value classification center data matrix with the current user feature data set and the current scenario feature data set, and input them into the user satisfaction level mapping model for mapping to obtain the current user satisfaction level data set.

9. The decision generation method based on big data for making color matching templates according to claim 8, characterized in that, The said S6 includes the following steps: S61. Use the row data in the sample feature value classification center data matrix corresponding to the current user satisfaction level data with the largest value in the current user satisfaction level data set as the current color adjustment strategy.

10. A system for implementing the decision-making generation method based on big data and applied to color palette template making as described in any one of claims 1-9.

Images