A method and system for non-branded commercial landscape renewal analysis

Abstract

The application relates to the technical field of landscape renewal analysis, and discloses a non-standard commercial landscape renewal analysis method and system, first case selection and data collection, second model construction and hyperparameter optimization, third model evaluation and result analysis, fourth SHAP analysis and result visualization, and fifth user interaction and feedback. The method can effectively evaluate the relative importance of spatial configuration, functional diversity and cultural symbolism, and provides a scientific basis and operation path for stock renewal of commercial real estate. In combination with the cross-case comparison method and the CatBoost algorithm, the spatial logic and practical path of non-standard commercial landscape renewal are systematically analyzed, and the shortcomings of existing research in quantitative analysis are made up. Through SHAP analysis, the importance of features is quantified, a scientific basis and operation path are provided for stock renewal of commercial real estate, and the transformation of commercial space from a functional interface to a public cultural field is promoted.

Description

Technical Field

[0001] This invention relates to the field of landscape analysis technology, specifically to a method and system for analyzing the renewal of non-standard commercial landscapes. Background Technology

[0002] With the acceleration of urbanization, commercial real estate has undergone a paradigm shift from incremental expansion to stock renewal. Statistics show that in 2024, the stock of shopping malls in China exceeded 6.3 × 10⁻⁶. 9 In third- and fourth-tier cities, the per capita commercial area has reached 1.7 square meters, far exceeding the international warning line (0.8-1.2 square meters). More than 30% of commercial properties face the dilemma of occupancy rates below 60%. Traditional standardized business models are no longer able to meet diversified consumer demands, leading to intensified homogeneous competition in commercial spaces and a decline in spatial vitality.

[0003] Although some progress has been made in the renewal of non-standard commercial landscapes, existing research still has the following shortcomings:

[0004] Theoretical blind spots: There has been no systematic analysis of the value leap of non-standard commercial spaces from homogeneous commercial functional landscapes to differentiated public cultural fields, and there is a lack of spatial practice, representation and representational space interaction coupling models based on Lefebvre's three-element framework.

[0005] Lack of universal patterns: Domestic research mostly focuses on traditional spaces, and foreign conclusions need to be verified locally, lacking the extraction of universal patterns across cases.

[0006] Methodological limitations: Existing studies mostly employ qualitative analysis and lack the support of quantitative models, making it difficult to accurately assess the relative importance of spatial configuration, functional diversity, and cultural symbolism. Summary of the Invention

[0007] The purpose of this invention is to solve the above-mentioned problems by designing a non-standard commercial landscape renewal analysis method and system.

[0008] The first aspect of this invention provides a method for analyzing the renewal of non-standard commercial landscapes, the method comprising the following steps:

[0009] The steps are: first, case selection and data collection; second, model building and hyperparameter tuning; third, model evaluation and result analysis; fourth, SHAP analysis and result visualization; and fifth, user interaction and feedback.

[0010] The first step, case selection and data collection, includes the logic for case selection, data collection methods, and data cleaning and preprocessing.

[0011] Case selection logic: Select representative non-standard commercial projects as research samples;

[0012] Data collection methods: Collect public perception data on the renewal of non-standard commercial landscapes through online and offline questionnaires and in-depth interviews;

[0013] Data cleaning and preprocessing: removing duplicate, invalid, and outlier data, and standardizing continuous variables;

[0014] The second step is model building and hyperparameter tuning, which includes initial model configuration, hyperparameter tuning, and model training.

[0015] Initial model configuration: Set the baseline parameters of the CatBoost model;

[0016] Hyperparameter tuning: A Bayesian optimization strategy combining five-fold cross-validation is used for hyperparameter tuning;

[0017] Model training: Retrain the CatBoost model using the optimal parameter combination;

[0018] The third step is model evaluation and results analysis, which includes the calculation of evaluation indicators and results analysis.

[0019] Evaluation index calculation: Calculate the mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and coefficient of determination (R²) on the test set.

[0020] Results Analysis: Analyze the model performance based on the evaluation index results;

[0021] Step 4: SHAP analysis and results visualization: This includes SHAP analysis, chart display, and report generation.

[0022] SHAP analysis: Calculate the SHAP value for each feature to quantify its contribution to the model's prediction results;

[0023] Charts show: a value transition path diagram of generated spatial representation, spatial practice, and representational space; and a cross-case comparison result diagram.

[0024] Report generation: Generate detailed analysis reports based on user needs;

[0025] Step 5: User Interaction and Feedback: This includes user interaction and feedback collection;

[0026] User interaction: Interacting with users through a graphical user interface (GUI) or a command-line interface (CLI);

[0027] Feedback collection: Collect user feedback on the analysis results to understand user needs and areas for improvement.

[0028] Optionally, in a first implementation of the first aspect of the present invention, the case selection logic is as follows: representative non-standard commercial projects are selected as research samples.

[0029] This includes selecting Guiyang Century Golden Resources Shopping Center, Kunming Century Golden Resources Shopping Center, and Xi'an Chang'an Fangyuanhui as research samples, representing three different regional cultural backgrounds: karst landform, Kunming's climate, and historical context. This ensures that the cases differ in regional culture, site constraints, and core customer groups, in order to extract differentiated practice paths.

[0030] Optionally, in a second implementation of the first aspect of the present invention, the 3. data collection method involves collecting public perception data on the updating of non-standard commercial landscapes through online and offline questionnaires and in-depth interviews;

[0031] The survey included online questionnaires distributed through various means such as social media and email, offline questionnaires conducted through intercept interviews in multiple activity venues in non-standard commercial districts and shopping malls, and in-depth interviews conducted with representative respondents to obtain more detailed descriptions of their spatial experiences.

[0032] Data cleaning and preprocessing: removing duplicate, invalid, and outlier data, and standardizing continuous variables;

[0033] This includes listing continuous variables systematically and labeling categorical variables to ensure data accuracy and consistency.

[0034] Optionally, in a third implementation of the first aspect of the present invention, step 4. Initial model configuration: setting the baseline parameters of the CatBoost model;

[0035] Baseline parameters include the number of iterations, learning rate, tree depth, and loss function.

[0036] Hyperparameter tuning: A Bayesian optimization strategy combining five-fold cross-validation is used for hyperparameter tuning;

[0037] This includes iteratively updating the prior distribution to efficiently explore the parameter space and find the optimal tree depth, number of iterations, L2 regularization coefficient, and learning rate parameters.

[0038] Model training: Retrain the CatBoost model using the optimal parameter combination;

[0039] The CatBoost model includes features to ensure that the model can fully learn the complex nonlinear relationships in the data.

[0040] Optionally, in the fourth implementation of the first aspect of the present invention, step 5. Evaluation index calculation: Calculate the mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and coefficient of determination (R²) evaluation indexes on the test set.

[0041] The model performance is analyzed in a concentrated manner using the calculated mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and coefficient of determination (R²).

[0042] Results Analysis: Analyze the model performance based on the evaluation index results;

[0043] The analysis of the indicator results ensures that the model can effectively capture the complex nonlinear relationships between spatial configuration, functional diversity, and cultural symbolism in the renewal of non-standard commercial landscapes. If the model performance does not meet the requirements, it is returned to the model building module for adjustment and optimization.

[0044] Optionally, in a fifth implementation of the first aspect of the present invention, step 6. SHAP analysis: calculating the SHAP value of each feature and quantifying its contribution to the model prediction results;

[0045] The SHAP beehive diagram and bar chart are used to visually display the ranking of feature importance, revealing the potential mechanisms of performance in non-standard commercial landscape renewal;

[0046] Charts show: a value transition path diagram of generated spatial representation, spatial practice, and representational space; and a cross-case comparison result diagram.

[0047] The spatial logic and practical path of non-standard commercial landscape renewal are intuitively displayed through value leap path diagrams and cross-case comparison result diagrams and data charts.

[0048] Report generation: Generate detailed analysis reports based on user needs;

[0049] The analysis report includes several main parts: case background, data collection and processing, model building and evaluation, SHAP analysis, and conclusions and recommendations.

[0050] Optionally, in a sixth implementation of the first aspect of the present invention, step 7, user interaction, involves interacting with the user through a graphical user interface (GUI) or a command-line interface (CLI).

[0051] On the interface above, users can select analysis cases, view model evaluation results, browse SHAP analysis charts, and export detailed analysis reports.

[0052] Feedback collection: Collect user feedback on the analysis results to understand user needs and areas for improvement;

[0053] After collecting user feedback, we continuously optimize system functions and analysis methods based on the feedback to improve system usability and user satisfaction.

[0054] The second aspect of the present invention provides a non-standard commercial landscape renewal analysis system, which includes: a data collection module: responsible for collecting public perception data on non-standard commercial landscape renewal, collecting data through online and offline questionnaires and in-depth interviews, and storing the collected data in a database for subsequent processing and analysis;

[0055] Data processing module: Cleans, standardizes, encodes features and balances the collected data. The z-score method is used to standardize continuous variables. The SMOTE synthetic minority oversampling technique is applied to balance the distribution of satisfaction scores. The dataset is divided into training and test sets in an 8:2 ratio. Five-fold cross-validation is performed on the training set to form the data partition.

[0056] Model building module: The regression model is built using the CatBoost algorithm to evaluate spatial configuration and functional diversity. CatBoost is an advanced ensemble learning algorithm based on the gradient boosting decision tree (GBDT) framework. It effectively handles heterogeneous data and reduces overfitting through ordered target encoding, symmetric tree structure, efficient gradient calculation and regularization techniques.

[0057] Model evaluation module: The model performance is evaluated using mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and coefficient of determination (R²). Each evaluation index is calculated on the test set, the model performance on different data is analyzed, and the model performance is analyzed based on the evaluation index results.

[0058] Results visualization module: The importance ranking of features is displayed intuitively through SHAP beehive diagrams and bar charts, revealing the potential mechanism of non-standard commercial landscape renewal performance. Furthermore, the spatial logic and practical path of non-standard commercial landscape renewal are displayed intuitively through chart data such as value leap path diagrams and cross-case comparison result diagrams.

[0059] User interaction module: Interacts with users through a graphical user interface (GUI) or a command-line interface (CLI), allowing users to input analysis requirements, view analysis results, and export reports.

[0060] This invention employs a cross-case comparison method combined with the CatBoost algorithm to systematically analyze the spatial logic and practical path of non-standard commercial landscape renewal. By collecting public perception data from multiple non-standard commercial projects, a CatBoost regression model is constructed to assess the relative importance of spatial configuration, functional diversity, and cultural symbolism on user satisfaction. SHAP analysis is used to interpret the model's prediction results, revealing the potential mechanisms underlying the performance of non-standard commercial landscape renewal. The system mainly includes a data collection module, a data processing module, a model building module, a model evaluation module, a results visualization module, and a user interaction module. These modules collaborate to complete the analysis task of non-standard commercial landscape renewal. Through the combination of cross-case comparison and machine learning algorithms, the spatial logic and practical path of non-standard commercial landscape renewal are systematically analyzed. Implementation cases demonstrate that this method can effectively assess the relative importance of spatial configuration, functional diversity, and cultural symbolism, providing a scientific basis and operational path for the renewal of existing commercial real estate. Future research can further expand the scope of analysis to examine the impact of technological innovation on the design and operation of non-standard commercial spaces, providing more precise guidance for the sustainable development of commercial real estate. Based on Lefebvre's theory of spatial production, this paper constructs a "breaking the circle" analysis model for the renewal of non-standard commercial landscapes, expanding the application boundaries of spatial production theory in the field of commercial landscapes. Combining cross-case comparison and the CatBoost algorithm, it systematically analyzes the spatial logic and practical path of non-standard commercial landscape renewal, making up for the shortcomings of existing research in quantitative analysis. Through SHAP analysis, the importance of characteristics is quantified, providing a scientific basis and operational path for the renewal of existing commercial real estate, promoting the transformation of commercial space from a functional interface to a public cultural field. Attached Figure Description

[0061] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention.

[0062] Figure 1 A flowchart of a ship traffic data security management method provided in an embodiment of the present invention;

[0063] Figure 2 This is a schematic diagram of the structure of the ship traffic data security management system provided in an embodiment of the present invention. Detailed Implementation

[0064] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” or “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, apparatus, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0065] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 A flowchart of a ship traffic data security management method provided in this embodiment of the invention is shown. The method specifically includes the following steps:

[0066] Step 101, Case Selection and Data Collection: This includes the logic for case selection, data collection methods, and data cleaning and preprocessing.

[0067] Case selection logic: Select representative non-standard commercial projects as research samples;

[0068] Data collection methods: Collect public perception data on the renewal of non-standard commercial landscapes through online and offline questionnaires and in-depth interviews;

[0069] Data cleaning and preprocessing: removing duplicate, invalid, and outlier data, and standardizing continuous variables;

[0070] In this embodiment, representative non-standard commercial projects are selected as research samples, including Guiyang Century Golden Resources Shopping Center, Kunming Century Golden Resources Shopping Center and Xi'an Chang'an Fangyuanhui, which respectively represent three different regional cultural backgrounds: karst landform, spring city climate and historical context. This ensures that the cases differ in regional culture, site constraints and core customer groups, so as to extract differentiated practice paths.

[0071] In this embodiment, the Guiyang project reinterprets the karst imagery with a youthful twist, constructing the core IP "Colorful Dopamine," simulating the light and shadow of karst caves to create a differentiated cognitive anchor. The Kunming project leverages its year-round spring-like climate to construct the "Free Flower Valley" IP system, creating an immersive experience through landscape elements such as three-dimensional flower borders, flowing flower islands, and terraced flower fields. The Xi'an project, centered on the dialogue between Chang'an's cultural heritage and contemporary life, uses "MIX Fangyuan Peninsula" as its core IP, abstractly translating historical symbols to achieve a contemporary adaptation of cultural representation.

[0072] In this embodiment, the questionnaire design involved developing a questionnaire to measure public perception of the renewal of non-standard commercial landscapes. Items were designed based on respondents' real experiences with the selected projects, and a five-point Likert scale (1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree) was used to indicate the degree of agreement. Data was collected through online and offline questionnaires, supplemented by in-depth interviews with members of the public who had visited at least one "renovated" non-standard commercial street within the past 12 months. After data cleaning, duplicate, invalid, and outlier data were removed, resulting in 353 valid responses, constituting the main dataset.

[0073] Step 102, Model Building and Hyperparameter Tuning: This includes initial model configuration, hyperparameter tuning, and model training;

[0074] Initial model configuration: Set the baseline parameters of the CatBoost model;

[0075] Hyperparameter tuning: A Bayesian optimization strategy combining five-fold cross-validation is used for hyperparameter tuning;

[0076] Model training: Retrain the CatBoost model using the optimal parameter combination;

[0077] In this embodiment, when setting the initial model configuration, the baseline parameters of the CatBoost model are set, including the number of iterations (iterations=1000), learning rate (learning_rate=0.1), tree depth (depth=6), and loss function (loss_function='RMSE').

[0078] In this embodiment, the parameter space is efficiently explored by continuously iterating and updating the prior distribution to find the optimal tree depth, number of iterations, L2 regularization coefficient, and learning rate parameters. A Bayesian optimization strategy combined with five-fold cross-validation is used for hyperparameter tuning to identify the optimal parameter combination that maximizes prediction accuracy, such as tree depth (depth=4), number of iterations (iterations=1674), L2 regularization coefficient (l2_leaf_reg=7), and learning rate (learning_rate=0.0393). Model training can then be performed by retraining the CatBoost model using the optimal parameter combination to ensure that the model can fully learn the complex nonlinear relationships in the data. The model performance is then validated on the test set to ensure the model's prediction accuracy, robustness, and interpretability.

[0079] Step 103, Model Evaluation and Result Analysis: This includes the calculation of evaluation indicators and result analysis;

[0080] Evaluation index calculation: Calculate the mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and coefficient of determination (R²) on the test set.

[0081] Results Analysis: Analyze the model performance based on the evaluation index results;

[0082] In this embodiment, the evaluation index calculation can assess the predictive performance of the model by calculating the mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and coefficient of determination (R²) on the test set. The results analysis then analyzes the model performance based on the evaluation index results to ensure that the model can effectively capture the complex nonlinear relationships between spatial configuration, functional diversity, and cultural symbolism in the renewal of non-standard commercial landscapes.

[0083] In this embodiment, mean absolute error (MAE), mean squared error (MSE), root mean square error (RMSE), and coefficient of determination (R²) are calculated on the test set to analyze model performance. If the model performance does not meet the requirements, the results are returned to the model building module for adjustment and optimization.

[0084] Step 104, SHAP Analysis and Results Visualization: This includes SHAP analysis, chart display, and report generation;

[0085] SHAP analysis: Calculate the SHAP value for each feature to quantify its contribution to the model's prediction results;

[0086] Charts show: a value transition path diagram of generated spatial representation, spatial practice, and representational space; and a cross-case comparison result diagram.

[0087] Report generation: Generate detailed analysis reports based on user needs;

[0088] In this embodiment, the SHAP value of each feature is calculated to quantify its contribution to the model's prediction results. The importance ranking of features is visually displayed using SHAP swarm diagrams and bar charts, revealing the potential mechanisms behind the performance of non-standard commercial landscape updates. A detailed analysis report is generated based on user needs, including case background, data collection and processing, model construction and evaluation, SHAP analysis, and conclusions and recommendations. SHAP analysis is used to explain the model's prediction results and quantify the contribution of each feature to user satisfaction. The analysis results are visually presented through charts and graphs, including spatial representation, the value transition path of spatial practice and reproducible space, and cross-case comparison results.

[0089] Step 105, User Interaction and Feedback: This includes user interaction and feedback collection;

[0090] User interaction: Interacting with users through a graphical user interface (GUI) or a command-line interface (CLI);

[0091] Feedback collection: Collect user feedback on the analysis results to understand user needs and areas for improvement.

[0092] In this embodiment, user interaction is achieved through a graphical user interface (GUI) or a command-line interface (CLI). Users can select analysis cases, view model evaluation results, browse SHAP analysis charts, and export detailed analysis reports. User feedback on the analysis results is collected to understand user needs and areas for improvement. Based on user feedback, system functions and analysis methods are continuously optimized to improve system usability and user satisfaction.

[0093] Please see Figure 2 A schematic diagram of the structure of the ship traffic data security management system provided in this embodiment of the invention. The system includes:

[0094] Data collection module: responsible for collecting public perception data on the renewal of non-standard commercial landscapes, collecting data through online and offline questionnaires and in-depth interviews, and storing the collected data in a database for subsequent processing and analysis;

[0095] Data processing module: Cleans, standardizes, encodes features and balances the collected data. The z-score method is used to standardize continuous variables. The SMOTE synthetic minority oversampling technique is applied to balance the distribution of satisfaction scores. The dataset is divided into training and test sets in an 8:2 ratio. Five-fold cross-validation is performed on the training set to form the data partition.

[0096] Model building module: The regression model is built using the CatBoost algorithm to evaluate spatial configuration and functional diversity. CatBoost is an advanced ensemble learning algorithm based on the gradient boosting decision tree (GBDT) framework. It effectively handles heterogeneous data and reduces overfitting through ordered target encoding, symmetric tree structure, efficient gradient calculation and regularization techniques.

[0097] Model evaluation module: The model performance is evaluated using mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and coefficient of determination (R²). Each evaluation index is calculated on the test set, the model performance on different data is analyzed, and the model performance is analyzed based on the evaluation index results.

[0098] Results visualization module: The importance ranking of features is displayed intuitively through SHAP beehive diagrams and bar charts, revealing the potential mechanism of non-standard commercial landscape renewal performance. Furthermore, the spatial logic and practical path of non-standard commercial landscape renewal are displayed intuitively through chart data such as value leap path diagrams and cross-case comparison result diagrams.

[0099] User interaction module: Interacts with users through a graphical user interface (GUI) or a command-line interface (CLI), allowing users to input analysis requirements, view analysis results, and export reports.

[0100] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described equipment or apparatus / unit can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0101] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0102] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for analyzing the renewal of non-standard commercial landscapes, characterized in that, The method includes the following steps: The steps are: first, case selection and data collection; second, model building and hyperparameter tuning; third, model evaluation and result analysis; fourth, SHAP analysis and result visualization; and fifth, user interaction and feedback. The first step, case selection and data collection, includes the logic for case selection, data collection methods, and data cleaning and preprocessing. Case selection logic: Select representative non-standard commercial projects as research samples; Data collection methods: Collect public perception data on the renewal of non-standard commercial landscapes through online and offline questionnaires and in-depth interviews; Data cleaning and preprocessing: removing duplicate, invalid, and outlier data, and standardizing continuous variables; The second step is model building and hyperparameter tuning, which includes initial model configuration, hyperparameter tuning, and model training. Initial model configuration: Set the baseline parameters of the CatBoost model; Hyperparameter tuning: A Bayesian optimization strategy combining five-fold cross-validation is used for hyperparameter tuning; Model training: Retrain the CatBoost model using the optimal parameter combination; The third step is model evaluation and results analysis, which includes the calculation of evaluation indicators and results analysis. Evaluation index calculation: Calculate the mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and coefficient of determination (R²) on the test set. Results Analysis: Analyze the model performance based on the evaluation index results; Step 4: SHAP analysis and results visualization: This includes SHAP analysis, chart display, and report generation. SHAP analysis: Calculate the SHAP value for each feature to quantify its contribution to the model's prediction results; Charts show: a value transition path diagram of generated spatial representation, spatial practice, and representational space; and a cross-case comparison result diagram. Report generation: Generate detailed analysis reports based on user needs; Step 5: User Interaction and Feedback: This includes user interaction and feedback collection; User interaction: Interacting with users through a graphical user interface (GUI) or a command-line interface (CLI); Feedback collection: Collect user feedback on the analysis results to understand user needs and areas for improvement.

2. The non-standard commercial landscape renewal analysis method as described in claim 1, characterized in that, The case selection logic is as follows: representative non-standard commercial projects are selected as research samples. This includes selecting Guiyang Century Golden Resources Shopping Center, Kunming Century Golden Resources Shopping Center, and Xi'an Chang'an Fangyuanhui as research samples, representing three different regional cultural backgrounds: karst landform, Kunming's climate, and historical context. This ensures that the cases differ in regional culture, site constraints, and core customer groups, in order to extract differentiated practice paths.

3. The non-standard commercial landscape renewal analysis method as described in claim 1, characterized in that, The data collection method involves collecting public perception data on the renewal of non-standard commercial landscapes through online and offline questionnaires and in-depth interviews. The survey included online questionnaires distributed through various means such as social media and email, offline questionnaires conducted through intercept interviews in multiple activity venues in non-standard commercial districts and shopping malls, and in-depth interviews conducted with representative respondents to obtain more detailed descriptions of their spatial experiences. Data cleaning and preprocessing: removing duplicate, invalid, and outlier data, and standardizing continuous variables; This includes listing continuous variables systematically and labeling categorical variables to ensure data accuracy and consistency.

4. The non-standard commercial landscape renewal analysis method as described in claim 1, characterized in that, The initial model configuration includes setting the baseline parameters of the CatBoost model. Baseline parameters include the number of iterations, learning rate, tree depth, and loss function; Hyperparameter tuning: A Bayesian optimization strategy combining five-fold cross-validation is used for hyperparameter tuning; This includes iteratively updating the prior distribution to efficiently explore the parameter space and find the optimal tree depth, number of iterations, L2 regularization coefficient, and learning rate parameters. Model training: Retrain the CatBoost model using the optimal parameter combination; The CatBoost model includes features to ensure that the model can fully learn the complex nonlinear relationships in the data.

5. The non-standard commercial landscape renewal analysis method as described in claim 1, characterized in that, The evaluation metrics are calculated as follows: mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and coefficient of determination (R²) on the test set. The model performance is analyzed in a concentrated manner using the calculated mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and coefficient of determination (R²). Results Analysis: Analyze the model performance based on the evaluation index results; The analysis of the indicator results ensures that the model can effectively capture the complex nonlinear relationships between spatial configuration, functional diversity, and cultural symbolism in the renewal of non-standard commercial landscapes. If the model performance does not meet the requirements, it is returned to the model building module for adjustment and optimization.

6. The non-standard commercial landscape renewal analysis method as described in claim 5, characterized in that, The SHAP analysis involves calculating the SHAP value for each feature to quantify its contribution to the model's prediction results. The SHAP beehive diagram and bar chart are used to visually display the ranking of feature importance, revealing the potential mechanisms of performance in non-standard commercial landscape renewal; Charts show: a value transition path diagram of generated spatial representation, spatial practice, and representational space; and a cross-case comparison result diagram. The spatial logic and practical path of non-standard commercial landscape renewal are intuitively displayed through value leap path diagrams and cross-case comparison result diagrams and data charts. Report generation: Generate detailed analysis reports based on user needs; The analysis report includes several main parts: case background, data collection and processing, model building and evaluation, SHAP analysis, and conclusions and recommendations.

7. The non-standard commercial landscape renewal analysis method as described in claim 1, characterized in that, The user interaction refers to interaction with the user through a graphical user interface (GUI) or a command-line interface (CLI). On the interface above, users can select analysis cases, view model evaluation results, browse SHAP analysis charts, and export detailed analysis reports. Feedback collection: Collect user feedback on the analysis results to understand user needs and areas for improvement; After collecting user feedback, we continuously optimize system functions and analysis methods based on the feedback to improve system usability and user satisfaction.

8. A non-standard commercial landscape renewal analysis system, characterized in that, The system includes: Data collection module: responsible for collecting public perception data on the renewal of non-standard commercial landscapes, collecting data through online and offline questionnaires and in-depth interviews, and storing the collected data in a database for subsequent processing and analysis; Data processing module: Cleans, standardizes, encodes features and balances the collected data. The z-score method is used to standardize continuous variables. The SMOTE synthetic minority oversampling technique is applied to balance the distribution of satisfaction scores. The dataset is divided into training and test sets in an 8:2 ratio. Five-fold cross-validation is performed on the training set to form the data partition. Model building module: The regression model is built using the CatBoost algorithm to evaluate spatial configuration and functional diversity. CatBoost is an advanced ensemble learning algorithm based on the gradient boosting decision tree (GBDT) framework. It effectively handles heterogeneous data and reduces overfitting through ordered target encoding, symmetric tree structure, efficient gradient calculation and regularization techniques. Model evaluation module: The model performance is evaluated using mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and coefficient of determination (R²). Each evaluation index is calculated on the test set, the model performance on different data is analyzed, and the model performance is analyzed based on the evaluation index results. Results visualization module: The importance ranking of features is displayed intuitively through SHAP beehive diagrams and bar charts, revealing the potential mechanism of non-standard commercial landscape renewal performance. Furthermore, the spatial logic and practical path of non-standard commercial landscape renewal are displayed intuitively through chart data such as value leap path diagrams and cross-case comparison result diagrams. User interaction module: Interacts with users through a graphical user interface (GUI) or a command-line interface (CLI), allowing users to input analysis requirements, view analysis results, and export reports.

Images