Enterprise digital management maturity evaluation method based on artificial intelligence

Abstract

The invention relates to the technical field of artificial intelligence, in particular to an artificial intelligence-based enterprise digital management maturity evaluation method, which comprises the following steps of: performing sample generation by adopting a generative adversarial network algorithm based on a momentum optimization variant to realize data expansion; inputting the expanded enterprise management data into a feature extraction model to train the feature extraction model; inputting the enterprise management data after feature extraction into a feature dimension reduction model to train the feature dimension reduction model; inputting the dimension-reduced enterprise management data into a classifier model to train a classifier; and processing new sample data by using the trained model so as to evaluate the digital management maturity of the new sample data.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an enterprise digital management maturity assessment method based on artificial intelligence. Background Art

[0002] With the rapid development of information technology, enterprises are gradually implementing digital transformation to improve operational efficiency and market competitiveness. Digital management not only involves the application of technology, but also includes the overall transformation of corporate culture, strategic decision-making and internal processes. Therefore, it is particularly important to accurately assess the maturity of enterprises in the process of digital transformation, which helps enterprises understand the current situation, formulate reasonable development strategies, and optimize management practices.

[0003] Existing technologies mainly rely on traditional data processing methods and statistical analysis to evaluate the digital management maturity of enterprises. These methods often rely on the subjective judgment of experts and simple quantitative analysis, and lack the ability to deeply mine and dynamically adapt to large-scale complex data. In addition, traditional methods have many limitations in data processing and model training, such as insufficient model generalization ability due to limited data volume, and serious information loss in feature extraction and dimensionality reduction.

[0004] These deficiencies affect the accuracy and efficiency of evaluation and limit the ability of enterprises to make timely and effective decisions in a rapidly changing market environment. Summary of the invention

[0005] The purpose of the present invention is to provide an enterprise digital management maturity assessment method based on artificial intelligence, which solves the problem that the shortcomings of existing enterprise digital management affect the accuracy and efficiency of the assessment and limit the ability of enterprises to make timely and effective decisions in a rapidly changing market environment.

[0006] To achieve the above object, the present invention provides an enterprise digital management maturity assessment method based on artificial intelligence, comprising the following steps:

[0007] Collect enterprise management data and mark it;

[0008] The generative adversarial network algorithm based on momentum optimization variant is used to generate samples and obtain enterprise management data expansion;

[0009] Input the expanded enterprise management data into the feature extraction model to train the feature extraction model;

[0010] Input the enterprise management data after feature extraction into the feature dimensionality reduction model to train the feature dimensionality reduction model;

[0011] Input the enterprise management data after dimension reduction into the classifier model to train the classifier;

[0012] Use the trained model to process new sample data to assess its digital management maturity.

[0013] Wherein, a generative adversarial network algorithm based on a momentum optimization variant is used to generate samples to obtain enterprise management data expansion, and the steps further include:

[0014] The generative adversarial network algorithm based on momentum optimization variant includes a generator and a discriminator, wherein the goal of the generator is to generate enterprise management data that is as realistic as possible to deceive the discriminator; the task of the discriminator is to distinguish between real enterprise management data and fake enterprise management data generated by the generator.

[0015] Among them, the training process of the generative adversarial network algorithm based on the momentum optimization variant includes:

[0016] Initialize the parameters of the GAN model based on a variant of momentum optimization;

[0017] The Adam optimizer with dynamic learning rate is used to optimize the parameters of the generator and discriminator;

[0018] In each training cycle, the generator generates a batch of fake enterprise management data, and the discriminator evaluates the difference between these enterprise management data and real enterprise management data. The feedback from the discriminator is used to adjust the parameters of the generator to make the generated enterprise management data more realistic.

[0019] Adaptive feature enhancement technology is used to optimize the quality of generated enterprise management data. By dynamically adjusting the key features in the generated enterprise management data, based on the feedback from the discriminator, the model’s adaptability and accuracy to complex enterprise management scenarios are enhanced.

[0020] According to the feedback from the discriminator, the strategy for generating enterprise management data is continuously adjusted to better adapt to the changes in the distribution of enterprise management data. The adjustment of the generation strategy is achieved by an adaptive mechanism.

[0021] After a fixed period, the generation effect of the model is evaluated and the parameters of the generator and discriminator are updated accordingly;

[0022] Repeat the above steps until the preset stop iteration condition is met, which means the model training is completed.

[0023] The expanded enterprise management data is input into the feature extraction model to train the feature extraction model, and the steps further include:

[0024] Initialize the neural network model;

[0025] In the growth stage, the fine-tuning of the acceleration weights can be used to quickly adapt to the characteristics of enterprise management data;

[0026] During the synthesis period, weight synthesis is performed;

[0027] In the evaluation period, the current network performance is evaluated and a decision is made whether to adjust the cycle optimization strategy. If yes, the cycle returns to the growth period; if no, the cycle proceeds to the pre-splitting period.

[0028] In prophase, which simulates the cell's preparation for division, the weights are slightly adjusted;

[0029] During the split period, structural adjustments were implemented;

[0030] During the dormant period, adjust the learning rate and decrease the learning rate;

[0031] Repeat the above steps until the preset stop iteration condition is met, which means the model training is completed.

[0032] The enterprise management data after feature extraction is input into the feature dimension reduction model to train the feature dimension reduction model, and the step further includes:

[0033] Initialize all weights of the autoencoder;

[0034] In the encoder training phase, the enterprise management data after feature extraction is input and converted into low-dimensional feature representation. Each input enterprise management data is gradually compressed through the encoder layers. Each layer of the encoder obtains a new output through the output of the previous layer and the weight calculation of the current layer;

[0035] After the features are encoded, the network weights are adjusted according to the inherent fuzziness of the enterprise management data. The importance and contribution of each feature are evaluated through fuzzy logic, and the weights are dynamically adjusted according to the evaluation results.

[0036] The decoder accepts the low-dimensional features output by the encoder and tries to reconstruct the original input data. Each low-dimensional feature is expanded layer by layer until the data with the same dimension as the original input is reconstructed.

[0037] By comparing the difference between the reconstructed data and the original input data, the error is calculated, and the weights in the network are updated through the back-propagation algorithm to reduce the reconstruction error;

[0038] Adopting an adaptive feature attenuation strategy, the feature dimension reduction quality and reconstruction accuracy of enterprise management data are optimized by dynamically adjusting the retention or attenuation degree of features in the decoding stage during the feature dimension reduction model training process of the autoencoder;

[0039] Repeat the above steps until the preset stop iteration condition is met, which means the model training is completed.

[0040] The enterprise management data after dimension reduction is input into the classifier model to train the classifier, and the steps further include:

[0041] In the initialization phase, the weights and biases from the input layer to the hidden layer are randomly initialized;

[0042] Calculate the representation of each sample in the hidden layer;

[0043] Use the least squares method to adjust the output layer weights and minimize the difference between the actual output and the expected output;

[0044] Fractional gradient adjustment is used to optimize weights by calculating fractional derivatives of gradients.

[0045] Repeat the above steps until the preset stop iteration condition is met, which means the model training is completed.

[0046] The enterprise digital management maturity assessment method based on artificial intelligence of the present invention adopts a generative adversarial network algorithm based on a momentum optimization variant for data expansion, and effectively generates realistic enterprise management data through the collaborative work of a generator and a discriminator. It can overcome the insufficient training problem caused by limited data samples, improve the generalization ability of the model, enhance the diversity and authenticity of the generated data, overcome the model overfitting problem caused by insufficient data volume, and improve the prediction accuracy of the model.

[0047] A circular dynamic optimization algorithm is used to optimize the parameters of the neural network and simulate the different stages of the cell cycle. By adjusting the optimization strategy in stages, the adaptability and accuracy of the model to complex data environments are enhanced, the risk of overfitting is effectively reduced, and the local optimal solution can be effectively escaped, thereby improving the accuracy and efficiency of feature extraction.

[0048] An autoencoder based on lightweight fuzzy weights is used to reduce the dimensionality of enterprise management data. By dynamically adjusting the fuzzy logic weights in the weight update process, the accuracy of the reduced dimensionality data and the generalization ability of the network are improved. By adjusting the fuzzy logic weights, the autoencoder not only reduces the reconstruction error, but also can more accurately reflect the structural characteristics of the original data, thereby enhancing the stability and reliability of data processing.

[0049] The extreme learning machine algorithm based on fractional gradient adjustment is used to improve the classifier training process. Through the careful control of fractional derivatives, the processing capability of nonlinear problems and the generalization of the classifier are optimized. Fractional gradient adjustment enables the extreme learning machine to adjust weights more finely during the learning process, reducing the risk of overfitting of the model and improving the ability to handle complex classification problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below.

[0051] Figure 1 It is a flow chart of the enterprise digital management maturity assessment method based on artificial intelligence according to the first embodiment of the present invention.

[0052] Figure 2 It is a training flow chart of a neural network algorithm based on circular dynamic optimization according to the first embodiment of the present invention.

[0053] Figure 3 It is a flowchart of model training and application according to the first embodiment of the present invention.

[0054] Figure 4 It is a step diagram of the enterprise digital management maturity assessment method based on artificial intelligence according to the first embodiment of the present invention. DETAILED DESCRIPTION

[0055] Embodiments of the present invention are described in detail below. Examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, but should not be construed as limiting the present invention.

[0056] The first embodiment of the present application is:

[0057] See also Figures 1 to 4 ,in, Figure 1 It is a flow chart of the enterprise digital management maturity assessment method based on artificial intelligence according to the first embodiment of the present invention. Figure 2 It is a training flow chart of a neural network algorithm based on circular dynamic optimization according to the first embodiment of the present invention. Figure 3 It is a flowchart of model training and application according to the first embodiment of the present invention. Figure 4 It is a step diagram of the enterprise digital management maturity assessment method based on artificial intelligence according to the first embodiment of the present invention.

[0058] The present invention provides an enterprise digital management maturity assessment method based on artificial intelligence, comprising the following steps:

[0059] S100: Collect enterprise management data and mark it;

[0060] Specifically, the data of the present invention comes from the internal management system of the enterprise, such as the ERP system, CRM system and human resource management platform, etc. The internal management system of the enterprise provides a large amount of original data generated by the enterprise during the operation process, and the collected data is stored in JSON format. In one embodiment, the attributes of the data include: ax1 is the total assets of the enterprise (numeric type); ax2 is the annual operating income (numeric type); ax3 is the total number of employees (numeric type); ax4 is the proportion of employees in the IT department (numeric type); ax5 is the annual IT investment amount (numeric type); ax6 is the service life of the ERP system (numeric type); ax7 is the service life of the CRM system (numeric type); ax8 is the cloud service adoption rate (percentage); ax9 is the data security investment ratio (percentage); ax10 is the customer satisfaction score (numeric type).

[0061] It should be noted that this embodiment is only used to illustrate one data format and type of the present invention. In actual applications, the attributes of enterprise management data are usually more than 10 attributes, and the number of attributes of the data may reach dozens or even hundreds.

[0062] Furthermore, the collected enterprise management data is labeled. The labeling method of the present invention is manual labeling. In one embodiment, the labeled categories include: "high maturity", "medium maturity" and "low maturity", a total of 3 categories.

[0063] S200: Generate samples using a generative adversarial network algorithm based on a momentum optimization variant to obtain enterprise management data augmentation;

[0064] Specifically, it can be understood that in the task of the present invention, the collection, acquisition, labeling and preprocessing of training data are time-consuming and labor-intensive, and insufficient training samples can easily lead to poor generalization ability of the model, while affecting the accuracy of the model. The present invention uses a generative adversarial network algorithm based on a momentum optimization variant to generate samples, thereby achieving enterprise management data expansion. The generative adversarial network algorithm based on a momentum optimization variant includes a generator (G) and a discriminator (D). The goal of the generator is to generate enterprise management data that is as realistic as possible to deceive the discriminator; the task of the discriminator is to distinguish between real enterprise management data and fake enterprise management data generated by the generator. The present invention uses a constrained optimization strategy in a traditional generative adversarial network to enhance the quality of generated enterprise management data, allowing the model to better adapt to various types of enterprise management data distributions, so that the generated enterprise management data has significant improvements in diversity and authenticity.

[0065] The training process of the Generative Adversarial Network algorithm based on the momentum optimization variant is as follows:

[0066] 1. Initialize the parameters of the generative adversarial network model based on the momentum optimization variant. In one embodiment, the initialization method uses a Gaussian distribution strategy for initialization, which is expressed as:

[0067]

[0068]

[0069] In the formula, represents the initial weight of the generator; represents the initial weight of the discriminator; and are the number of nodes in the input layer and the output layer respectively; is the Gaussian distribution function.

[0070] 2. Use the Adam optimizer with dynamic learning rate to optimize the parameters of the generator and discriminator. The optimization method is:

[0071]

[0072]

[0073] In the formula, and Represent the parameters of the generator and the discriminator respectively; It is The learning rate of the iteration; and are the first and second moment estimates; is a small positive number added to prevent division by zero. Preferably, Set to 0.0001; The simulated annealing strategy is used for setting, that is, as the number of iterations increases, the learning rate decreases at a rate of 0.95.

[0074] Furthermore, the first-order moment estimate and the second moment estimate The calculation method is expressed as:

[0075]

[0076]

[0077] In the formula, and are the first-order and second-order moment estimates of the previous iteration, respectively; It is in The gradient of the iteration; and is the decay rate parameter. Preferably, and Both are set to 0.85.

[0078] 3. In each training cycle, the generator generates a batch of fake enterprise management data, and the discriminator evaluates the difference between these enterprise management data and real enterprise management data. The feedback of the discriminator is used to adjust the parameters of the generator to make the generated enterprise management data more realistic. Specifically, the method of generating enterprise management data and evaluating it is expressed as:

[0079]

[0080]

[0081]

[0082] In the formula, is the output of the discriminator, is the generated enterprise management data after feature enhancement, indicating the probability of judging it as real enterprise management data; The generator is based on the input and parameters Generated enterprise management data; is the function of the discriminator; and They are the loss functions of the generator and the discriminator respectively; It is a label. The real enterprise management data is 1, and the generated enterprise management data is 0.

[0083] 4. Adopt adaptive feature enhancement technology to optimize the quality of generated enterprise management data. By dynamically adjusting the key features in the generated enterprise management data, according to the feedback of the discriminator, the adaptability and accuracy of the model to complex enterprise management scenarios are enhanced. Specifically, define the feature difference evaluation function The function to quantify the key feature differences between generated enterprise management data and real enterprise management data is calculated as follows:

[0084]

[0085] In the formula, and Represent the feature vectors of generated enterprise management data and real enterprise management data respectively; is the number of features; For the The importance weight of each feature; : Generate the kth feature of the feature vector of enterprise management data, : The kth feature of the feature vector of real enterprise management data.

[0086] Furthermore, based on the feature difference evaluation, an adaptive feature enhancement strategy function is defined The calculation method is expressed as:

[0087]

[0088] In the formula, is the learning rate; Indicates the generation of enterprise management data The gradient of the feature difference evaluation function. Preferably, Set to 0.01.

[0089] In one embodiment, the feature importance weights The update strategy is expressed as:

[0090]

[0091] In the formula, is a smoothing parameter. Preferably, Set to 0.5.

[0092] 5. According to the feedback from the discriminator, the strategy for generating enterprise management data is continuously adjusted to better adapt to the changes in the distribution of enterprise management data. The adjustment of the generation strategy is implemented by an adaptive mechanism, which is expressed as:

[0093]

[0094]

[0095] In the formula, is the dynamic adjustment factor of the generator learning rate; is the learning step size; is the adjustment amount of the parameter.

[0096] Furthermore, the learning rate dynamic adjustment factor The calculation method is expressed as:

[0097]

[0098] In the formula, is the decay rate of the adjustment factor; It is in Iterations for parameters The amount of adjustment; represents the second norm of the adjustment amount. Preferably, Set to 0.98.

[0099] 6. After a fixed period, evaluate the generation effect of the model and update the parameters of the generator and discriminator accordingly. The method of updating the parameters is expressed as:

[0100]

[0101]

[0102] In the formula, is the first momentum factor; is the second momentum factor; is the updated generator parameter; is the generator parameter before updating; is the updated discriminator parameter; is the discriminator parameter before updating. Preferably, the fixed period is set to every 10 iterations.

[0103] In one embodiment, the first momentum factor and the second momentum factor The calculation method is expressed as:

[0104]

[0105]

[0106] In the formula, and is the initial momentum factor; and is the rate at which the momentum factor decays with the number of iterations; is the current step number of the training iteration. Preferably, Set to 0.9, Set to 0.8.

[0107] 7. Repeat the above steps until the preset stop iteration condition is met, indicating that the model training is completed. In one embodiment, the preset stop iteration condition is reaching a preset maximum number of iterations. Preferably, the preset maximum number of iterations is set to 1000 times.

[0108] S300: inputting the expanded enterprise management data into the feature extraction model to train the feature extraction model;

[0109] Specifically, the expanded enterprise management data is input into the feature extraction model for training. The traditional gradient descent method relies on global gradient information to update weights. It is inspired by the periodic metabolic process of biological cells and optimizes the response and adaptability of organisms through different mechanisms at different stages.

[0110] The present invention uses a circular dynamic optimization algorithm to optimize the parameters of the neural network. The circular dynamic optimization algorithm adjusts the optimization strategy at different training stages to simulate the various stages in the cell cycle. Different strategies are used in each stage to adjust the weights and biases of the neural network, so that the algorithm can more flexibly adapt to changes in the management data of the training enterprise, reduce the risk of overfitting, and effectively jump out of the local optimal solution.

[0111] The training process of the neural network algorithm based on ring dynamic optimization is as follows:

[0112] 1. Initialize the neural network model. The present invention adopts a 3-layer fully connected neural network for feature extraction. The number of hidden layers of the neural network is 2. The first hidden layer includes 100 neurons, the second hidden layer includes 50 neurons, and the activation function is the ReL_RU activation function.

[0113] Furthermore, initialize the weights of the neural network and bias , and set the initial learning rate In one embodiment, the initial weights and biases are set as follows:

[0114]

[0115]

[0116] In the formula, It means that it obeys a specific distribution. The mean is 0 and the variance is The normal distribution of is the variance of the initial distribution. Preferably, Set to 0.01.

[0117] 2. During the growth stage, the algorithm quickly adapts to the characteristics of enterprise management data by fine-tuning the acceleration weights.

[0118] The present invention adopts the accelerated gradient descent algorithm to update the neural network parameters, and the update method is expressed as:

[0119]

[0120]

[0121] In the formula, is the acceleration factor, is the loss function for neural network training, is the gradient of the loss function with respect to the weight parameter, is the gradient of the loss function with respect to the bias parameter, For the The weight parameter of the iteration, For the The bias parameter for the iteration, For the The weight parameter of the iteration, For the The bias parameter for the iteration, is the symbol of partial derivative, is the learning rate during the growth phase. Preferably, the loss function Using cross entropy loss, Set to 0.01.

[0122] Furthermore, the acceleration factor According to the decline rate and curvature of the loss function, in order to avoid over-adjustment, the calculation method is:

[0123]

[0124] In the formula, is the preset maximum acceleration factor, is the minimum value function.

[0125] 3. In the synthesis phase, the algorithm performs weight synthesis, that is, synthesizes multiple weight parameters into new parameters through specific operations to improve the generalization ability of the network. In one embodiment, the synthesis function is defined as The calculation method is:

[0126]

[0127] In the formula, is the weight The importance coefficient of is the number of weight parameters.

[0128] Furthermore, the importance coefficient Based on the previous stage The gradient size is calculated dynamically. The calculation method of the importance coefficient is expressed as:

[0129]

[0130] In the formula, is the decay rate parameter.

[0131] Furthermore, the decay rate parameter The calculation method of control is:

[0132]

[0133] 4. During the evaluation period, the algorithm evaluates the current network performance and decides whether to adjust the periodic optimization strategy. In one embodiment, the performance evaluation function is set , outputs whether to enter the next cycle in the form of a decision function, and the decision method is expressed as:

[0134]

[0135] In the formula, is the performance threshold. Preferably, Set to 0.5. If the calculation is 1, it enters the next cycle, that is, the growth phase; otherwise it enters the pre-mitotic phase.

[0136] 5. In the prophase of mitosis, simulating the stage of cells preparing to divide, the weights are slightly adjusted. In one embodiment, a perturbation method is used. and As the adjustment increment of weights and biases, the calculation method is expressed as:

[0137]

[0138]

[0139] In the formula, is the adjustment increment of the weight parameter, is the adjustment increment of the bias parameter, and are the amplitude and frequency parameters for fine tuning. Preferably, Set to 5, Set to 3.14.

[0140] Further, and The setting method is expressed as:

[0141]

[0142]

[0143] In the formula, is a scaling factor for the perturbation amplitude. Preferably, Set to 0.95.

[0144] 6. During the splitting phase, the algorithm implements a major structural adjustment. In one embodiment, the adjustment is to use weight pruning to constrain the weights. The weight pruning function The calculation method is expressed as:

[0145]

[0146] In the formula, is the gradient threshold. Preferably, Set to 0.01, that is, only parameters with weights greater than 0.01 are retained.

[0147] 7. During the dormant period, the network gradually reaches a stable state based on the existing weights. At this stage, the learning rate is adjusted , so that the learning rate decreases, and the adjustment method is expressed as:

[0148]

[0149] In the formula, is the attenuation coefficient, is the number of completed cycles. Preferably, Set to 0.95.

[0150] 8. Repeat the above steps until the preset stop iteration condition is met, indicating that the model training is completed. In one embodiment, the preset stop iteration condition is reaching a preset maximum number of iterations. Preferably, the preset maximum number of iterations is set to 1000 times.

[0151] S400: inputting the enterprise management data after feature extraction into the feature dimension reduction model to train the feature dimension reduction model;

[0152] Specifically, the enterprise management data after feature extraction is input into the feature dimensionality reduction model to train the feature dimensionality reduction model. The present invention adopts an autoencoder based on lightweight fuzzy weights as a feature dimensionality reduction algorithm. The autoencoder compresses the input enterprise management data into a low-dimensional space by learning and reconstructs it back to the original space, thereby achieving dimensionality reduction of the enterprise management data. In addition, the autoencoder includes an encoder part and a decoder part. The encoder is responsible for compressing the input enterprise management data into low-dimensional features, and the decoder attempts to reconstruct the original input from these low-dimensional features. The present invention improves the autoencoder, adopts fuzzy logic weights to adjust the weight update process in the network, and dynamically adjusts based on the characteristics of the input enterprise management data, which can better handle the uncertainty and ambiguity in the enterprise management data, thereby improving the accuracy of the dimensionality reduction process and the generalization ability of the network.

[0153] The training process of the autoencoder algorithm based on lightweight fuzzy weights is as follows:

[0154] 1. Initialize all weights of the autoencoder. In one embodiment, a small random number initialization method is used to break the symmetry and avoid the gradient vanishing problem. The initialization method is expressed as:

[0155]

[0156] In the formula, Represents the first weights; is the standard deviation. Preferably, Set to 0.01.

[0157] 2. In the encoder training stage, the enterprise management data after input feature extraction is converted into low-dimensional feature representation. Each input enterprise management data is gradually compressed through the encoder layer. Each layer of the encoder obtains a new output through the output of the previous layer and the weight calculation of the current layer. Each layer of the encoder outputs The calculation method is expressed as:

[0158]

[0159] In the formula, It is The output of the layer; is the input vector or the output of the previous layer; Is the connection The input and The weight of the output; It is Bias of the layer; is the ReLU activation function.

[0160] 3. After the features are encoded, the network weights are adjusted according to the inherent fuzziness of the enterprise management data. The importance and contribution of each feature are evaluated through fuzzy logic. The weights are dynamically adjusted according to these evaluation results. The adjustment method is expressed as:

[0161]

[0162]

[0163] In the formula, is the adjusted weight parameter; is the amount of weight adjustment; is a weight adjustment factor based on fuzzy logic; is the learning rate. Preferably, Set to 0.01.

[0164] Furthermore, the weight adjustment factor According to the input and output The relationship is dynamically calculated, and the calculation method is expressed as:

[0165]

[0166] In the formula, is the data input to the i-th layer of the autoencoder; is the output of the i-th layer of the autoencoder; is the fuzziness factor. Preferably, Set to 0.1.

[0167] 4. The decoder accepts the low-dimensional features output by the encoder and then tries to reconstruct the original input data. Each low-dimensional feature is expanded layer by layer until the data with the same dimension as the original input is reconstructed. The output of the decoder is The calculation method is expressed as:

[0168]

[0169] In the formula, is the output of the reconstruction; is the weight in the decoder; is the bias in the decoder.

[0170] 5. By comparing the difference between the reconstructed data and the original input data, the error is calculated, and the weights in the network are updated through the back propagation algorithm to reduce the reconstruction error. The update method of the gradient of and weight is expressed as:

[0171]

[0172]

[0173] In the formula, is the reconstruction error; is the original input data; is the output of the decoder; is the derivative of the ReLU activation function.

[0174] Furthermore, the gradient The calculation method is expressed as:

[0175]

[0176]

[0177] In the formula, The decoder is The weighted sum of the inputs to the neurons; is the output of the encoder, which is the input of the decoder; is the activation function in The derivative at .

[0178] 6. Adopting the adaptive feature attenuation strategy, the feature attenuation model training process of the autoencoder dynamically adjusts the retention or attenuation degree of the feature in the decoding stage to optimize the dimensionality reduction quality and reconstruction accuracy of the enterprise management data. Specifically, define the feature attenuation function Used to adjust the decoder output , expressed as:

[0179]

[0180] In the formula, is the raw output of the decoder; is the decay rate parameter; is the decaying relevance scoring function.

[0181] In one embodiment, the calculation method of the decay correlation score function is expressed as:

[0182]

[0183] In the formula, is the reconstruction error Output to decoder The gradient of is the feature vector output by the decoder.

[0184] In this embodiment, the attenuation rate is dynamically adjusted. The adjustment method is expressed as:

[0185]

[0186] In the formula, is the smoothing factor; yes The variance of yes Preferably, Set to 0.5.

[0187] 7. Repeat the above steps until the preset stop iteration condition is met, indicating that the model training is completed. In one embodiment, the preset stop iteration condition is reaching the preset maximum number of iterations. Preferably, the preset maximum number of iterations is set to 1000. After the model training is completed, the obtained That is, the enterprise management data features after feature dimensionality reduction.

[0188] S500: inputting the enterprise management data after dimension reduction into the classifier model to train the classifier;

[0189] Specifically, the enterprise management data after dimensionality reduction is input into the classifier model to train the classifier. The present invention adopts an extreme learning machine algorithm based on fractional gradient adjustment as the classification algorithm, and improves the extreme learning machine. Specifically, a fractional gradient adjustment mechanism is adopted based on the fractional calculus theory to perform more precise control on the updating of weights in the learning process of the classifier, thereby improving the model's processing capability and generalization of nonlinear problems.

[0190] The training process of the extreme learning machine algorithm based on fractional gradient adjustment is as follows:

[0191] 1. In the initialization stage, the weights from the input layer to the hidden layer and bias Perform random initialization. In one embodiment, the initialization method is expressed as:

[0192]

[0193]

[0194] In the formula, is the standard deviation used for weights and biases during initialization. Preferably, Set to 0.1.

[0195] 2. Calculate the representation of each sample in the hidden layer and the output of the hidden layer By input data With randomly generated weights and bias After the linear combination of Convert to:

[0196]

[0197] In the formula, is the input data; is the weight of the hidden layer; is the bias of the hidden layer; is the Sigmoid activation function.

[0198] 3. Use the least squares method to adjust the output layer weights , to minimize the actual output and expected output The difference between them is calculated as:

[0199]

[0200] In the formula, is the transpose of the hidden layer output; is the regularization parameter; is the target output; is the identity matrix, and The same dimensions. Preferably, Set to a matrix of all 1s.

[0201] 4. Fractional gradient adjustment is used to optimize the weights by calculating the fractional derivative of the gradient, so as to reduce the risk of overfitting while maintaining the learning speed. The update method of the output layer weight is expressed as:

[0202]

[0203]

[0204] In the formula, is the updated output layer weight; is the update increment of the output layer weight; is the error function; Express The fractional derivative of ; is the learning rate; is a fractional order parameter.

[0205] Furthermore, the learning rate The dynamic learning rate adjustment mechanism is used to dynamically adjust the learning rate by real-time analysis of the error change rate during training, thereby optimizing the convergence speed and classification accuracy of the model. Specifically, the learning rate The adjustment is based on the error function The current gradient and the previous gradient The rate of change of is expressed as follows:

[0206]

[0207] In the formula, To adjust the updated learning rate; is the learning rate adjustment factor, usually set to 0.1; The function is a symbolic function; and are the error gradients of the current and previous iterations, respectively.

[0208] Furthermore, the calculation method of the error function is expressed as:

[0209]

[0210] In one embodiment, the fractional derivative Characterization of the error function The application of fractional calculus provides more detailed weight adjustment than traditional gradient, and the calculation method is expressed as:

[0211]

[0212] In the formula, is the gamma function, which extends the traditional factorial to the real and complex domains; is the fractional order parameter, defining the non-integer number of derivative operations; is the error function About Output No. Preferably, the fractional order parameter Set to 0.5.

[0213] 5. Repeat the above steps until the preset stop iteration condition is met, indicating that the model training is completed. In one embodiment, the preset stop iteration condition is reaching a preset maximum number of iterations. Preferably, the preset maximum number of iterations is set to 1200 times.

[0214] S600: Use the trained model to process new sample data to assess its digital management maturity.

[0215] Specifically, the trained model is used to process new sample data to evaluate the digital management maturity of the data. In one embodiment, there is a new sample data set for an enterprise. The new sample is input into a trained feature extraction model for feature extraction. Furthermore, the data after feature extraction is input into a trained data dimension reduction model for data dimension reduction. Furthermore, the reduced features are input into a trained classifier model to classify the digital management maturity of the enterprise according to the features. In this embodiment, the classification categories include: "high maturity", "medium maturity" and "low maturity", a total of 3 categories. Based on this, the process of model training and application of the present invention is as follows: Figure 3 shown.

[0216] A generative adversarial network algorithm based on a momentum optimization variant is used for data augmentation. Through the collaborative work of the generator and the discriminator, realistic enterprise management data can be effectively generated. This can overcome the insufficient training problem caused by limited data samples, improve the generalization ability of the model, enhance the diversity and authenticity of the generated data, overcome the model overfitting problem caused by insufficient data, and improve the prediction accuracy of the model.

[0217] A circular dynamic optimization algorithm is used to optimize the parameters of the neural network and simulate the different stages of the cell cycle. By adjusting the optimization strategy in stages, the adaptability and accuracy of the model to complex data environments are enhanced, the risk of overfitting is effectively reduced, and the local optimal solution can be effectively escaped, thereby improving the accuracy and efficiency of feature extraction.

[0218] An autoencoder based on lightweight fuzzy weights is used to reduce the dimensionality of enterprise management data. By dynamically adjusting the fuzzy logic weights in the weight update process, the accuracy of the reduced dimensionality data and the generalization ability of the network are improved. By adjusting the fuzzy logic weights, the autoencoder not only reduces the reconstruction error, but also can more accurately reflect the structural characteristics of the original data, thereby enhancing the stability and reliability of data processing.

[0219] The extreme learning machine algorithm based on fractional gradient adjustment is used to improve the classifier training process. Through the careful control of fractional derivatives, the processing capability of nonlinear problems and the generalization of the classifier are optimized. Fractional gradient adjustment enables the extreme learning machine to adjust weights more finely during the learning process, reducing the risk of overfitting of the model and improving the ability to handle complex classification problems.

[0220] What is disclosed above is only one or more preferred embodiments of the present application, and cannot be used to limit the scope of rights of the present application. Ordinary technicians in this field can understand that all or part of the processes of implementing the above embodiments and equivalent changes made according to the claims of the present application are still within the scope covered by the present application.

Claims

1. An enterprise digital management maturity assessment method based on artificial intelligence, characterized in that: The following steps are involved: Collect enterprise management data and mark it; The generative adversarial network algorithm based on momentum optimization variant is used to generate samples and obtain enterprise management data expansion; Input the expanded enterprise management data into the feature extraction model to train the feature extraction model; Input the enterprise management data after feature extraction into the feature dimensionality reduction model to train the feature dimensionality reduction model; Input the enterprise management data after dimension reduction into the classifier model to train the classifier; Use the trained model to process new sample data to assess the digital management maturity of the enterprise.

2. The enterprise digital management maturity assessment method based on artificial intelligence according to claim 1, characterized in that: A generative adversarial network algorithm based on a momentum optimization variant is used to generate samples to obtain enterprise management data expansion. The step also includes: the generative adversarial network algorithm based on a momentum optimization variant includes a generator and a discriminator, wherein the goal of the generator is to generate realistic enterprise management data to deceive the discriminator; the task of the discriminator is to distinguish between real enterprise management data and fake enterprise management data generated by the generator.

3. The enterprise digital management maturity assessment method based on artificial intelligence as claimed in claim 2, characterized in that: The training process of the GAN algorithm based on the momentum optimization variant includes: Initialize the parameters of the generative adversarial network model based on the momentum optimization variant, where the initialization method uses the Gaussian distribution strategy for initialization, expressed as: , In the formula, represents the initial weight of the generator; represents the initial weight of the discriminator; and are the number of nodes in the input layer and the output layer respectively; is a Gaussian distribution function; the Adam optimizer with dynamic learning rate is used to optimize the parameters of the generator and discriminator. The optimization method is: , In the formula, and Represent the parameters of the generator and the discriminator respectively; It is The learning rate of the iteration; and are the first and second moment estimates; This is to prevent small positive numbers added by division by zero; in each training cycle, the generator generates a batch of fake enterprise management data, and the discriminator evaluates the difference between these enterprise management data and real enterprise management data. The feedback of the discriminator is used to adjust the parameters of the generator to make the enterprise management data it generates more realistic. The method of generating enterprise management data and evaluating it is expressed as: , , In the formula, is the output of the discriminator, is the generated enterprise management data after feature enhancement, indicating the probability of judging it as real enterprise management data; The generator is based on the input and parameters Generated enterprise management data; is the function of the discriminator; and They are the loss functions of the generator and the discriminator respectively; is a label, the real enterprise management data is 1, and the generated enterprise management data is 0; the adaptive feature enhancement technology is used to optimize the quality of the generated enterprise management data, and the key features in the generated enterprise management data are dynamically adjusted according to the feedback of the discriminator to enhance the adaptability and accuracy of the model to complex enterprise management scenarios. Among them, the feature difference evaluation function is defined The function to quantify the key feature differences between generated enterprise management data and real enterprise management data is calculated as follows: In the formula, and Represent the feature vectors of generated enterprise management data and real enterprise management data respectively; is the number of features; For the Importance weights of features, based on feature difference evaluation, define adaptive feature enhancement strategy function The calculation method is expressed as: In the formula, is the learning rate; Indicates the generation of enterprise management data The gradient of the feature difference evaluation function; : Generate the kth feature of the feature vector of enterprise management data, : The kth feature of the feature vector of real enterprise management data; According to the feedback from the discriminator, the strategy for generating enterprise management data is continuously adjusted to adapt to the changes in the distribution of enterprise management data. The adjustment of the generation strategy is implemented by an adaptive mechanism, which is expressed as: , In the formula, is the dynamic adjustment factor of the generator learning rate; is the learning step size; is the parameter adjustment amount, the learning rate dynamic adjustment factor The calculation method is expressed as: In the formula, is the decay rate of the adjustment factor; It is in Iterations for parameters The amount of adjustment; Represents the second norm of the adjustment amount; after a fixed period, the generation effect of the model is evaluated, and the parameters of the generator and discriminator are updated accordingly. The method of updating the parameters is expressed as: , In the formula, is the first momentum factor; is the second momentum factor; is the updated generator parameter; is the generator parameter before updating; is the updated discriminator parameter; is the discriminator parameter before updating; Repeat the above steps until the preset stop iteration condition is met, which means the model training is completed.

4. The enterprise digital management maturity assessment method based on artificial intelligence as claimed in claim 3 is characterized in that: Inputting the expanded enterprise management data into the feature extraction model to train the feature extraction model, the step further includes: Initialize the neural network model; In the growth stage, the fine-tuning of the acceleration weights can be used to quickly adapt to the characteristics of enterprise management data; During the synthesis period, weight synthesis is performed; In the evaluation period, the current network performance is evaluated and a decision is made whether to adjust the cycle optimization strategy. If yes, the cycle returns to the growth period; if no, the cycle proceeds to the pre-splitting period. In prophase, which simulates the cell's preparation for division, the weights are slightly adjusted; During the split period, structural adjustments were implemented; During the dormant period, adjust the learning rate and decrease the learning rate; Repeat the above steps until the preset stop iteration condition is met, which means the model training is completed.

5. The enterprise digital management maturity assessment method based on artificial intelligence as claimed in claim 4 is characterized in that: Inputting the enterprise management data after feature extraction into the feature dimensionality reduction model to train the feature dimensionality reduction model, the step further includes: Initialize all weights of the autoencoder; In the encoder training phase, the enterprise management data after feature extraction is input and converted into low-dimensional feature representation. Each input enterprise management data is gradually compressed through the encoder layers. Each layer of the encoder obtains a new output through the output of the previous layer and the weight calculation of the current layer; After the features are encoded, the network weights are adjusted according to the inherent fuzziness of the enterprise management data. The importance and contribution of each feature are evaluated through fuzzy logic, and the weights are dynamically adjusted according to the evaluation results. The decoder accepts the low-dimensional features output by the encoder and tries to reconstruct the original input data. Each low-dimensional feature is expanded layer by layer until the data with the same dimension as the original input is reconstructed. By comparing the difference between the reconstructed data and the original input data, the error is calculated, and the weights in the network are updated through the back-propagation algorithm to reduce the reconstruction error; Adopting an adaptive feature attenuation strategy, the feature dimension reduction quality and reconstruction accuracy of enterprise management data are optimized by dynamically adjusting the retention or attenuation degree of features in the decoding stage during the feature dimension reduction model training process of the autoencoder; Repeat the above steps until the preset stop iteration condition is met, which means the model training is completed.

6. The enterprise digital management maturity assessment method based on artificial intelligence according to claim 5, characterized in that: Inputting the enterprise management data after dimension reduction into the classifier model to train the classifier, the steps further include: In the initialization phase, the weights and biases from the input layer to the hidden layer are randomly initialized; Calculate the representation of each sample in the hidden layer; Use the least squares method to adjust the output layer weights and minimize the difference between the actual output and the expected output; Fractional gradient adjustment is used to optimize weights by calculating fractional derivatives of gradients. Repeat the above steps until the preset stop iteration condition is met, which means the model training is completed.

7. The enterprise digital management maturity assessment method based on artificial intelligence according to claim 5, characterized in that: In the encoder training stage, the enterprise management data after feature extraction is input and converted into a low-dimensional feature representation. Each input enterprise management data is gradually compressed through the encoder layer. Each layer of the encoder obtains a new output through the output of the previous layer and the weight calculation of the current layer. The steps also include: Each layer of the encoder outputs The calculation method is expressed as: In the formula, It is The output of the layer; is the input vector or the output of the previous layer; Is the connection The input and The weight of the output; It is Bias of the layer; is the ReLU activation function.

8. The enterprise digital management maturity assessment method based on artificial intelligence as claimed in claim 5, characterized in that: After the features are encoded, the network weights are adjusted according to the inherent fuzziness of the enterprise management data, the importance and contribution of each feature are evaluated through fuzzy logic, and the weights are dynamically adjusted according to the evaluation results. The steps also include: The adjustment method is expressed as: , In the formula, is the adjusted weight parameter; is the amount of weight adjustment; is a weight adjustment factor based on fuzzy logic; is the learning rate, the weight adjustment factor According to the input and output The relationship is dynamically calculated, and the calculation method is expressed as: In the formula, is the data input to the i-th layer of the autoencoder; is the output of the i-th layer of the autoencoder; is the fuzziness factor.

9. The enterprise digital management maturity assessment method based on artificial intelligence as claimed in claim 5, characterized in that: The decoder accepts the low-dimensional features output by the encoder and attempts to reconstruct the original input data. Each low-dimensional feature is expanded layer by layer until data with the same dimension as the original input is reconstructed. The steps also include: the output of the decoder The calculation method is expressed as: In the formula, is the output of the reconstruction; is the weight in the decoder; is the bias in the decoder.

10. The enterprise digital management maturity assessment method based on artificial intelligence according to claim 5, characterized in that: By comparing the difference between the reconstructed data and the original input data, calculating the error, and updating the weights in the network through the back propagation algorithm to reduce the reconstruction error, the step also includes: error The update method of the gradient of and weight is expressed as: , In the formula, is the reconstruction error; is the original input data; is the output of the decoder; is the derivative of the ReLU activation function, the gradient The calculation method is expressed as: , In the formula, The decoder is The weighted sum of the inputs to the neurons; is the output of the encoder, which is the input of the decoder; is the activation function in The derivative at .

Images