Scientific and technological innovation project risk processing method and device based on artificial intelligence, computer equipment and storage medium

Abstract

The invention relates to a scientific and technological innovation project risk processing method and device based on artificial intelligence, computer equipment and a storage medium. The method comprises the following steps: preprocessing project data of a scientific and technological innovation project to obtain preprocessed project data; screening out a target risk prediction model matched with the data features of the preprocessed project data; inputting the feature vector of the preprocessed project data into a target risk prediction model to obtain a current risk value of the scientific and technological innovation project; determining a current risk level of the scientific and technological innovation project according to the current risk value; obtaining a future risk level of the scientific and technological innovation project, and performing fusion processing on the current risk level and the future risk level to obtain a target risk level of the scientific and technological innovation project; and generating a current risk processing instruction corresponding to the target risk level, and performing corresponding risk processing on the scientific and technological innovation project according to the current risk processing instruction. By adopting the method, the risk processing accuracy of the scientific and technological innovation project can be improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for risk management of scientific and technological innovation projects based on artificial intelligence. Background Art

[0002] At present, in order to ensure the smooth operation of scientific and technological innovation projects, it is crucial to accurately handle the risks of scientific and technological innovation projects.

[0003] In traditional technology, when handling risks in scientific and technological innovation projects, manual experience and simple rules are generally used; however, this method is highly subjective, and the judgment standards of different evaluators vary, resulting in low accuracy in risk management of scientific and technological innovation projects. Summary of the Invention

[0004] Based on this, it is necessary to provide an artificial intelligence-based scientific and technological innovation project risk management method, device, computer equipment, computer-readable storage medium and computer program product to address the above technical problems, which can improve the risk management accuracy of scientific and technological innovation projects.

[0005] In the first aspect, this application provides a method for managing the risks of scientific and technological innovation projects based on artificial intelligence, including:

[0006] Obtaining project data of the scientific and technological innovation project to be analyzed; the project data at least includes project background information, project technical indicator data, market data, and project team information of the scientific and technological innovation project;

[0007] Preprocessing the project data to obtain preprocessed project data of the scientific and technological innovation project;

[0008] Determining data features of the preprocessed project data, and screening a target risk prediction model that matches the data features from a plurality of trained risk prediction models;

[0009] Performing feature extraction processing on the pre-processed project data to obtain a feature vector of the pre-processed project data, and inputting the feature vector into the target risk prediction model to obtain a current risk value of the scientific and technological innovation project;

[0010] Determining the current risk level of the scientific and technological innovation project based on the current risk value;

[0011] Obtaining a historical risk level of the scientific and technological innovation project, and determining a future risk level of the scientific and technological innovation project based on the historical risk level and the current risk level;

[0012] Fusing the current risk level and the future risk level to obtain a target risk level for the scientific and technological innovation project;

[0013] Generate a current risk handling instruction corresponding to the target risk level, and perform corresponding risk handling on the scientific and technological innovation project according to the current risk handling instruction.

[0014] In one embodiment, the pre-processed project data at least includes pre-processed project background information, pre-processed project technical indicator data, pre-processed market data and pre-processed project team information of the scientific and technological innovation project;

[0015] The performing feature extraction processing on the preprocessed project data to obtain a feature vector of the preprocessed project data includes:

[0016] performing feature extraction processing on the preprocessed project background information, the preprocessed project technical indicator data, the preprocessed market data, and the preprocessed project team information, respectively, to obtain a first feature vector of the preprocessed project background information, a second feature vector of the preprocessed project technical indicator data, a third feature vector of the preprocessed market data, and a fourth feature vector of the preprocessed project team information;

[0017] The first eigenvector, the second eigenvector, the third eigenvector and the fourth eigenvector are fused to obtain a eigenvector of the preprocessed project data.

[0018] In one embodiment, preprocessing the project data to obtain preprocessed project data of the scientific and technological innovation project includes:

[0019] Determining a hash value of each sub-item data in the item data, and determining duplicate data in the item data based on the hash value of each sub-item data;

[0020] Deleting duplicate data from the project data to obtain deduplicated project data;

[0021] The deduplicated project data is normalized to obtain normalized project data, which serves as preprocessed project data for the scientific and technological innovation project.

[0022] In one embodiment, determining the current risk level of the scientific and technological innovation project based on the current risk value includes:

[0023] When the current risk value is less than a first set threshold, determining that the current risk level of the scientific and technological innovation project is a low risk level;

[0024] When the current risk value is greater than or equal to the first set threshold and less than or equal to the second set threshold, determining that the current risk level of the scientific and technological innovation project is a medium risk level;

[0025] When the current risk value is greater than the second set threshold, it is determined that the current risk level of the scientific and technological innovation project is a high risk level.

[0026] In one embodiment, each trained risk prediction model is trained in the following manner:

[0027] Obtain sample project data of sample scientific and technological innovation projects;

[0028] Preprocessing the sample project data to obtain preprocessed sample project data of the sample scientific and technological innovation project;

[0029] Performing feature extraction processing on the preprocessed sample project data to obtain a sample feature vector of the preprocessed sample project data, and inputting the sample feature vector into a risk prediction model to be trained to obtain a predicted risk value of the sample scientific and technological innovation project;

[0030] The actual risk value of the sample scientific and technological innovation project is obtained, and according to the difference between the predicted risk value and the actual risk value, the risk prediction model to be trained is iteratively trained to obtain the trained risk prediction model.

[0031] In one embodiment, the iterative training of the risk prediction model to be trained based on the difference between the predicted risk value and the actual risk value to obtain the trained risk prediction model includes:

[0032] Determining a prediction error value of the risk prediction model to be trained based on a difference between the predicted risk value and the actual risk value;

[0033] When the prediction error value is greater than or equal to the error value threshold, adjusting the model parameters of the risk prediction model to be trained according to the prediction error value to obtain a risk prediction model after the model parameters are adjusted;

[0034] The risk prediction model after the model parameters are adjusted is trained again until the prediction error value obtained by the trained risk prediction model is less than the error value threshold, then the training is stopped, and the trained risk prediction model is used as the trained risk prediction model.

[0035] Secondly, this application also provides an artificial intelligence-based risk management device for scientific and technological innovation projects, including:

[0036] A data acquisition module is used to acquire project data of the scientific and technological innovation project to be analyzed; the project data at least includes project background information, project technical indicator data, market data and project team information of the scientific and technological innovation project;

[0037] A data processing module, configured to pre-process the project data to obtain pre-processed project data of the scientific and technological innovation project;

[0038] A model screening module is used to determine the data characteristics of the pre-processed project data and screen out a target risk prediction model that matches the data characteristics from multiple trained risk prediction models;

[0039] a risk prediction module, configured to perform feature extraction processing on the pre-processed project data to obtain a feature vector of the pre-processed project data, and input the feature vector into the target risk prediction model to obtain a current risk value of the scientific and technological innovation project;

[0040] A first determination module is configured to determine a current risk level of the scientific and technological innovation project based on the current risk value;

[0041] A second determination module is configured to obtain a historical risk level of the scientific and technological innovation project and determine a future risk level of the scientific and technological innovation project based on the historical risk level and the current risk level;

[0042] A level fusion module, configured to fuse the current risk level and the future risk level to obtain a target risk level for the scientific and technological innovation project;

[0043] The risk processing module is used to generate a current risk processing instruction corresponding to the target risk level, and perform corresponding risk processing on the scientific and technological innovation project according to the current risk processing instruction.

[0044] In a third aspect, the present application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0045] Obtaining project data of the scientific and technological innovation project to be analyzed; the project data at least includes project background information, project technical indicator data, market data, and project team information of the scientific and technological innovation project;

[0046] Preprocessing the project data to obtain preprocessed project data of the scientific and technological innovation project;

[0047] Determining data features of the preprocessed project data, and screening a target risk prediction model that matches the data features from a plurality of trained risk prediction models;

[0048] Performing feature extraction processing on the pre-processed project data to obtain a feature vector of the pre-processed project data, and inputting the feature vector into the target risk prediction model to obtain a current risk value of the scientific and technological innovation project;

[0049] Determining the current risk level of the scientific and technological innovation project based on the current risk value;

[0050] Obtaining a historical risk level of the scientific and technological innovation project, and determining a future risk level of the scientific and technological innovation project based on the historical risk level and the current risk level;

[0051] Fusing the current risk level and the future risk level to obtain a target risk level for the scientific and technological innovation project;

[0052] Generate a current risk handling instruction corresponding to the target risk level, and perform corresponding risk handling on the scientific and technological innovation project according to the current risk handling instruction.

[0053] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0054] Obtaining project data of the scientific and technological innovation project to be analyzed; the project data at least includes project background information, project technical indicator data, market data, and project team information of the scientific and technological innovation project;

[0055] Preprocessing the project data to obtain preprocessed project data of the scientific and technological innovation project;

[0056] Determining data features of the preprocessed project data, and screening a target risk prediction model that matches the data features from a plurality of trained risk prediction models;

[0057] Performing feature extraction processing on the pre-processed project data to obtain a feature vector of the pre-processed project data, and inputting the feature vector into the target risk prediction model to obtain a current risk value of the scientific and technological innovation project;

[0058] Determining the current risk level of the scientific and technological innovation project based on the current risk value;

[0059] Obtaining a historical risk level of the scientific and technological innovation project, and determining a future risk level of the scientific and technological innovation project based on the historical risk level and the current risk level;

[0060] Fusing the current risk level and the future risk level to obtain a target risk level for the scientific and technological innovation project;

[0061] Generate a current risk handling instruction corresponding to the target risk level, and perform corresponding risk handling on the scientific and technological innovation project according to the current risk handling instruction.

[0062] In a fifth aspect, the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the following steps:

[0063] Obtaining project data of the scientific and technological innovation project to be analyzed; the project data at least includes project background information, project technical indicator data, market data, and project team information of the scientific and technological innovation project;

[0064] Preprocessing the project data to obtain preprocessed project data of the scientific and technological innovation project;

[0065] Determining data features of the preprocessed project data, and screening a target risk prediction model that matches the data features from a plurality of trained risk prediction models;

[0066] Performing feature extraction processing on the pre-processed project data to obtain a feature vector of the pre-processed project data, and inputting the feature vector into the target risk prediction model to obtain a current risk value of the scientific and technological innovation project;

[0067] Determining the current risk level of the scientific and technological innovation project based on the current risk value;

[0068] Obtaining a historical risk level of the scientific and technological innovation project, and determining a future risk level of the scientific and technological innovation project based on the historical risk level and the current risk level;

[0069] Fusing the current risk level and the future risk level to obtain a target risk level for the scientific and technological innovation project;

[0070] Generate a current risk handling instruction corresponding to the target risk level, and perform corresponding risk handling on the scientific and technological innovation project according to the current risk handling instruction.

[0071] The above-mentioned artificial intelligence-based scientific and technological innovation project risk processing method, device, computer equipment, storage medium and computer program product first obtains the project background information, project technical indicator data, market data and project team information of the scientific and technological innovation project to be analyzed as the project data of the scientific and technological innovation project, and preprocesses the project data to obtain preprocessed project data of the scientific and technological innovation project, then determines the data characteristics of the preprocessed project data, and screens out a target risk prediction model that matches the data characteristics from multiple trained risk prediction models. Then, feature extraction processing is performed on the preprocessed project data to obtain a feature vector of the preprocessed project data, and the feature vector is input into the target risk prediction model to obtain the current risk value of the scientific and technological innovation project, and the current risk level of the scientific and technological innovation project is determined based on the current risk value. Then, the historical risk level of the scientific and technological innovation project is obtained, and the future risk level of the scientific and technological innovation project is determined based on the historical risk level and the current risk level. The current risk level and the future risk level are then integrated to obtain the target risk level of the scientific and technological innovation project. Finally, a current risk processing instruction corresponding to the target risk level is generated, and the corresponding risk processing is performed on the scientific and technological innovation project according to the current risk processing instruction. In this way, in the process of risk management of scientific and technological innovation projects, the comprehensiveness of the risk assessment basis can be ensured through multi-dimensional data collection, the data quality can be improved with the help of data preprocessing, and the most suitable risk prediction model can be matched based on data features to enhance the model's pertinence. The current risk can be accurately quantified through feature extraction and model calculation, and dynamic trend analysis can be constructed in combination with historical risk levels. The current future risk level and the future risk level can be integrated to take into account the immediate status and development situation. Finally, targeted processing instructions are generated based on the integrated target risk level, which is conducive to improving the accuracy of risk management of scientific and technological innovation projects. Moreover, the entire process does not require human intervention, avoiding the high subjectivity in the method of using manual experience and simple rules for processing, and the differences in the judgment standards of different evaluators, which leads to the defect of low risk management accuracy of scientific and technological innovation projects, thereby improving the risk management accuracy of scientific and technological innovation projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments of the present application or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying any creative work.

[0073] Figure 1 A flowchart of a method for handling risks of scientific and technological innovation projects based on artificial intelligence in one embodiment;

[0074] Figure 2 A flowchart of a method for handling risks of scientific and technological innovation projects based on artificial intelligence in another embodiment;

[0075] Figure 3 Schematic diagram of a process for predicting risk of scientific and technological innovation projects based on artificial intelligence in one embodiment;

[0076] Figure 4 This is a structural block diagram of a risk management device for scientific and technological innovation projects based on artificial intelligence in one embodiment;

[0077] Figure 5 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0078] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0079] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0080] In an exemplary embodiment, Figure 1 As shown, a method for handling risks of scientific and technological innovation projects based on artificial intelligence is provided. This embodiment uses the method applied to a server as an example for illustration; it is understandable that the method can also be applied to a terminal, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. The terminal can be, but is not limited to, various personal computers, laptops, smartphones, and tablet computers; the server can be implemented as an independent server or a server cluster consisting of multiple servers. In this embodiment, the method includes the following steps:

[0081] Step S101: Acquire project data of a scientific and technological innovation project to be analyzed; the project data at least includes project background information, project technical indicator data, market data, and project team information of the scientific and technological innovation project.

[0082] Among them, scientific and technological innovation projects refer to projects carried out around new technologies, new methods or new applications, which can be power projects.

[0083] Among them, project background information includes the original intention of the project, technology research and development direction and other information of the scientific and technological innovation project.

[0084] Among them, project technical indicator data include technical innovation indicator data and technical maturity indicator data of scientific and technological innovation projects.

[0085] Among them, market data includes data such as the target market size, market growth rate, and competitor situation of scientific and technological innovation projects.

[0086] Among them, project team information includes professional background, project experience and other information of team members of scientific and technological innovation projects.

[0087] Exemplarily, the server obtains the research field information and research cycle information of the candidate scientific and technological innovation projects; then, the server selects the scientific and technological innovation projects whose research field information meets the preset research field information and whose research cycle information is greater than the preset research cycle information from the candidate scientific and technological innovation projects, as the scientific and technological innovation projects to be analyzed; then, the server obtains the project background information, project technical indicator data, market data and project team information of the scientific and technological innovation projects, all of which are used as the project data of the scientific and technological innovation projects to be analyzed.

[0088] Step S102: pre-process the project data to obtain pre-processed project data of the scientific and technological innovation project.

[0089] The preprocessed project data refers to the project data after preprocessing.

[0090] Exemplarily, the server identifies the noise information of the project data and determines the denoising method corresponding to the project data according to the noise information; then, the server denoises the project data according to the denoising method corresponding to the project data to obtain the pre-processed project data of the scientific and technological innovation project.

[0091] Step S103: determine the data features of the pre-processed project data, and select a target risk prediction model that matches the data features from a plurality of trained risk prediction models.

[0092] Among them, data features are used to represent attribute information extracted from a set of variables used to describe the characteristics of scientific and technological innovation projects.

[0093] Among them, the risk prediction model refers to a network model that can predict the risks of scientific and technological innovation projects, such as the CNN (Convolutional Neural Network) model.

[0094] Among them, the target risk prediction model refers to the risk prediction model that matches the data characteristics.

[0095] Exemplarily, the server extracts the data features of the preprocessed project data according to a preset feature extraction method; then, the server selects a risk prediction model that matches the data features from multiple trained risk prediction models as the target risk prediction model; for example, when the data features are linearly separable, the server uses the logistic regression algorithm as the target risk prediction model; when the data features are nonlinear boundary features, the server uses the vector machine algorithm as the target risk prediction model; when the data features have high feature dimensions and there is a lot of noise, the server uses the random forest algorithm as the target risk prediction model.

[0096] Step S104: perform feature extraction processing on the pre-processed project data to obtain a feature vector of the pre-processed project data, and input the feature vector into the target risk prediction model to obtain the current risk value of the scientific and technological innovation project.

[0097] Among them, the feature vector is used to represent the representation vector of the preprocessed project data.

[0098] Among them, the current risk value is used to represent the quantitative risk assessment results of scientific and technological innovation projects, and the corresponding value range can be [0,100].

[0099] Exemplarily, the server determines the data type of the preprocessed project data, and queries the correspondence between the data type and the feature extraction model, and obtains the feature extraction model corresponding to the data type of the preprocessed project data as the target feature extraction model corresponding to the preprocessed project data; then, the server inputs the preprocessed project data into the target feature extraction model for feature extraction processing to obtain a feature vector of the preprocessed project data; then, the server inputs the feature vector into the target risk prediction model to obtain a first risk value of the scientific and technological innovation project, and inputs the feature vector into the historical risk prediction model corresponding to the target risk prediction model to obtain a second risk value of the scientific and technological innovation project; then, the server sums the first risk value and the second risk value according to the first weight corresponding to the target risk prediction model and the second weight corresponding to the historical risk prediction model to obtain the current risk value of the scientific and technological innovation project.

[0100] Step S105: Determine the current risk level of the scientific and technological innovation project based on the current risk value.

[0101] Among them, the current risk level is used to indicate the risk level classification results of scientific and technological innovation projects at the current time, including high risk level, medium risk level and low risk level.

[0102] Exemplarily, the server queries the correspondence between the risk value and the risk level, and obtains the risk level corresponding to the current risk value as the current risk level of the scientific and technological innovation project.

[0103] Step S106: Obtain the historical risk level of the scientific and technological innovation project, and determine the future risk level of the scientific and technological innovation project based on the historical risk level and the current risk level.

[0104] Among them, the historical risk level is used to indicate the risk level classification results of scientific and technological innovation projects in historical time, including high risk level, medium risk level and low risk level.

[0105] Among them, the future risk level is used to indicate the risk level classification results of scientific and technological innovation projects in the future, including high risk level, medium risk level and low risk level.

[0106] Exemplarily, the server obtains the project identification of the scientific and technological innovation project, and based on the project identification, determines the historical risk level corresponding to the project identification from the database as the historical risk level of the scientific and technological innovation project; then, the server determines the risk level trend chart corresponding to the scientific and technological innovation project based on the historical risk level and the current risk level; then, the server determines the future risk level of the scientific and technological innovation project based on the risk level trend chart.

[0107] Step S107: The current risk level and the future risk level are integrated to obtain the target risk level of the scientific and technological innovation project.

[0108] Among them, the target risk level is used to represent the final risk level classification results of scientific and technological innovation projects, including high risk level, medium risk level and low risk level.

[0109] For example, the server fuses the current risk level and the future risk level according to a preset fusion rule table to obtain the target risk level of the scientific and technological innovation project; the server can also use the higher risk level between the current risk level and the future risk level as the target risk level of the scientific and technological innovation project.

[0110] Step S108: Generate a current risk handling instruction corresponding to the target risk level, and perform corresponding risk handling on the scientific and technological innovation project according to the current risk handling instruction.

[0111] Among them, the current risk handling instruction refers to the risk handling instruction corresponding to the target risk level.

[0112] Exemplarily, the server inputs the target risk level into the trained risk handling instruction prediction model to obtain the predicted probability of the target risk level under each preset risk handling instruction; then, the server selects the preset risk handling instruction with the largest predicted probability from each preset risk handling instruction as the current risk handling instruction corresponding to the target risk level; then, the server performs a rationality check on the current risk handling instruction to obtain a verification result of the current risk handling instruction; when the verification result indicates that the current risk handling instruction has passed the verification, the server performs corresponding risk handling on the scientific and technological innovation project in accordance with the current risk handling instruction.

[0113] In the above-mentioned artificial intelligence-based risk management method for scientific and technological innovation projects, the project background information, project technical indicator data, market data and project team information of the scientific and technological innovation project to be analyzed are first obtained as the project data of the scientific and technological innovation project, and the project data is preprocessed to obtain the preprocessed project data of the scientific and technological innovation project. Then, the data characteristics of the preprocessed project data are determined, and a target risk prediction model that matches the data characteristics is screened out from multiple trained risk prediction models. Then, feature extraction processing is performed on the preprocessed project data to obtain a feature vector of the preprocessed project data, and the feature vector is input into the target risk prediction model to obtain the current risk value of the scientific and technological innovation project, and the current risk level of the scientific and technological innovation project is determined based on the current risk value. Then, the historical risk level of the scientific and technological innovation project is obtained, and the future risk level of the scientific and technological innovation project is determined based on the historical risk level and the current risk level. Then, the current risk level and the future risk level are fused to obtain the target risk level of the scientific and technological innovation project. Finally, a current risk handling instruction corresponding to the target risk level is generated, and the corresponding risk handling of the scientific and technological innovation project is performed according to the current risk handling instruction. In this way, in the process of risk management of scientific and technological innovation projects, the comprehensiveness of the risk assessment basis can be ensured through multi-dimensional data collection, the data quality can be improved with the help of data preprocessing, and the most suitable risk prediction model can be matched based on data features to enhance the model's pertinence. The current risk can be accurately quantified through feature extraction and model calculation, and dynamic trend analysis can be constructed in combination with historical risk levels. The current future risk level and the future risk level can be integrated to take into account the immediate status and development situation. Finally, targeted processing instructions are generated based on the integrated target risk level, which is conducive to improving the accuracy of risk management of scientific and technological innovation projects. Moreover, the entire process does not require human intervention, avoiding the high subjectivity in the method of using manual experience and simple rules for processing, and the differences in the judgment standards of different evaluators, which leads to the defect of low risk management accuracy of scientific and technological innovation projects, thereby improving the risk management accuracy of scientific and technological innovation projects.

[0114] In an exemplary embodiment, the pre-processed project data includes at least pre-processed project background information, pre-processed project technical indicator data, pre-processed market data and pre-processed project team information of the scientific and technological innovation project.

[0115] Then, the above-mentioned step S104 performs feature extraction processing on the preprocessed project data to obtain the feature vector of the preprocessed project data, which specifically includes the following contents: performing feature extraction processing on the preprocessed project background information, the preprocessed project technical indicator data, the preprocessed market data and the preprocessed project team information respectively to obtain the first feature vector of the preprocessed project background information, the second feature vector of the preprocessed project technical indicator data, the third feature vector of the preprocessed market data, and the fourth feature vector of the preprocessed project team information; and performing fusion processing on the first feature vector, the second feature vector, the third feature vector and the fourth feature vector to obtain the feature vector of the preprocessed project data.

[0116] The pre-processed project background information refers to the project background information after pre-processing.

[0117] The pre-processed project technical indicator data refers to the project technical indicator data after pre-processing.

[0118] The pre-processed market data refers to the market data after pre-processing.

[0119] The pre-processed project team information refers to the project team information after pre-processing.

[0120] The first eigenvector is used to represent the representation vector of the project background information after preprocessing.

[0121] The second eigenvector is used to represent the characterization vector of the project technical indicator data after preprocessing.

[0122] The third eigenvector is used to represent the representation vector of the pre-processed market data.

[0123] The fourth eigenvector is used to represent the representation vector of the project team information after preprocessing.

[0124] Exemplarily, the server uses the pre-processed project background information as the main data, and the pre-processed project technical indicator data, pre-processed market data and pre-processed project team information as auxiliary data, and inputs them into the feature extraction model for feature extraction processing to obtain the first feature vector of the pre-processed project background information; then, the server uses the pre-processed project technical indicator data as the main data, and the pre-processed project background information, pre-processed market data and pre-processed project team information as auxiliary data, and inputs them into the feature extraction model for feature extraction processing to obtain the second feature vector of the pre-processed project technical indicator data; then, the server uses the pre-processed market data as the main data, and the pre-processed project background information, pre-processed market data and pre-processed project team information as auxiliary data, and inputs them into the feature extraction model for feature extraction processing to obtain the second feature vector of the pre-processed project technical indicator data. The server uses the preprocessed project team information as the main data and the preprocessed project background information, the preprocessed project technical indicator data and the preprocessed market data as auxiliary data, and inputs them into the feature extraction model for feature extraction processing to obtain the third eigenvector of the preprocessed project team information; then, the server uses the preprocessed project team information as the main data and the preprocessed project background information, the preprocessed project technical indicator data and the preprocessed market data as auxiliary data, and inputs them into the feature extraction model for feature extraction processing to obtain the fourth eigenvector of the preprocessed project team information; then, the server sums the first eigenvector, the second eigenvector, the third eigenvector and the fourth eigenvector according to their respective weights to obtain the eigenvector of the preprocessed project data.

[0125] In this embodiment, by extracting features from the multi-dimensional data in the pre-processed project data separately and then fusing them to generate an overall feature vector, the core information of the data in each dimension can be retained in a targeted manner, so that the obtained feature vector can more comprehensively characterize the essence of project risk, thereby improving the accuracy and robustness of the subsequent risk prediction model, and providing a more scientific decision-making basis for risk level classification and processing strategy generation.

[0126] In an exemplary embodiment, the above step S102 pre-processes the project data to obtain pre-processed project data of the scientific and technological innovation project, which specifically includes the following contents: determining the hash value of each sub-project data in the project data, and determining the duplicate data in the project data based on the hash value of each sub-project data; deleting the duplicate data in the project data to obtain deduplicated project data; normalizing the deduplicated project data to obtain normalized project data as pre-processed project data of the scientific and technological innovation project.

[0127] The sub-project data is used to represent independent sub-data units obtained by splitting the project data.

[0128] The hash value refers to a string of fixed length generated by performing hash calculation on the sub-project data.

[0129] Duplicate data refers to sub-project data with exactly the same hash value in the project data.

[0130] The deduplicated project data refers to the project data obtained by deleting duplicate data.

[0131] The normalized project data refers to the deduplicated project data after normalization.

[0132] Exemplarily, the server performs hash calculation on each sub-project data in the project data through a preset hash function to obtain a hash value of each sub-project data in the project data; then, the server determines the sub-project data in the project data with exactly the same hash value based on the hash value of each sub-project data, as duplicate data in the project data; then, the server deletes the duplicate data in the project data to obtain deduplicated project data; then, the server normalizes the deduplicated project data through the Z-Score (standard score) normalization formula to obtain normalized project data, and uses the normalized project data as preprocessed project data for the scientific and technological innovation project.

[0133] In this embodiment, by accurately locating and deleting duplicate sub-project data in project data through hash values, information redundancy can be effectively eliminated, data purity can be improved, and misjudgment of risk indicators due to data duplication can be avoided; at the same time, the deduplicated data is normalized, and data of different dimensions and different value ranges can be uniformly mapped to the standard interval, eliminating the interference of scale differences on subsequent feature extraction and risk model training.

[0134] In an exemplary embodiment, the above step S105 determines the current risk level of the scientific and technological innovation project based on the current risk value, specifically including the following contents: when the current risk value is less than the first set threshold, the current risk level of the scientific and technological innovation project is determined to be a low risk level; when the current risk value is greater than or equal to the first set threshold and less than or equal to the second set threshold, the current risk level of the scientific and technological innovation project is determined to be a medium risk level; when the current risk value is greater than the second set threshold, the current risk level of the scientific and technological innovation project is determined to be a high risk level.

[0135] The first set threshold refers to the critical value used to divide the low risk level into the medium risk level.

[0136] The second set threshold refers to the critical value used to divide the high risk level into the medium risk level.

[0137] Exemplarily, the server obtains the historical risk values ​​of historical scientific and technological innovation projects corresponding to the scientific and technological innovation projects, and obtains the mean and standard deviation of the historical risk values; then, the server obtains a first set threshold value based on the mean and standard deviation of the historical risk values ​​and a first preset multiple (for example, 0.5); for example, the product of the standard deviation of the historical risk values ​​and the first preset multiple is added to the mean of the historical risk values ​​to obtain the first set threshold value; then, the server obtains a first set threshold value based on the mean and standard deviation of the historical risk values ​​and a second preset multiple (for example, 1.5); for example, the product of the standard deviation of the historical risk values ​​and the second preset multiple is added to the mean of the historical risk values ​​to obtain the second set threshold value; then, the server uses the first set threshold value and the second set threshold value to judge the current risk value; when the current risk value is less than the first set threshold value, the server determines that the current risk level of the scientific and technological innovation project is a low risk level; when the current risk value is greater than or equal to the first set threshold value and less than or equal to the second set threshold value, the server determines that the current risk level of the scientific and technological innovation project is a medium risk level; when the current risk value is greater than the second set threshold value, the server determines that the current risk level of the scientific and technological innovation project is a high risk level.

[0138] In this embodiment, the current risk value is divided into intervals by setting a first threshold and a second threshold, and the continuous risk value is converted into discrete and clear risk levels, making the project risk assessment results more intuitive and easy to understand; at the same time, the standardized grading rules effectively avoid the subjectivity of risk assessment and ensure the consistency and comparability of the assessment results.

[0139] In an exemplary embodiment, the artificial intelligence-based scientific and technological innovation project risk management method provided in the present application also includes a training process for each trained risk prediction model, which specifically includes the following contents: obtaining sample project data of a sample scientific and technological innovation project; preprocessing the sample project data to obtain preprocessed sample project data of the sample scientific and technological innovation project; performing feature extraction processing on the preprocessed sample project data to obtain a sample feature vector of the preprocessed sample project data, and inputting the sample feature vector into the risk prediction model to be trained to obtain a predicted risk value of the sample scientific and technological innovation project; obtaining the actual risk value of the sample scientific and technological innovation project, and iteratively training the risk prediction model to be trained based on the difference between the predicted risk value and the actual risk value to obtain a trained risk prediction model.

[0140] Among them, the sample scientific and technological innovation projects refer to scientific and technological innovation projects used to train the risk prediction model to be trained.

[0141] Among them, sample project data refers to the project data of sample scientific and technological innovation projects.

[0142] The pre-processed sample project data refers to the sample project data after pre-processing.

[0143] The sample feature vector is used to represent the representation vector of the sample item data after preprocessing.

[0144] Among them, the predicted risk value refers to the predicted result corresponding to the risk value of the sample scientific and technological innovation project.

[0145] Among them, the actual risk value refers to the actual result corresponding to the risk value of the sample scientific and technological innovation project.

[0146] Exemplarily, the server obtains the research field information and research cycle information of the scientific and technological innovation projects to be analyzed, and selects the scientific and technological innovation projects to be analyzed whose research field information and research cycle information are the same as the research field information and research cycle information of the scientific and technological innovation projects from each scientific and technological innovation project to be analyzed as sample scientific and technological innovation projects; then, the server obtains sample project data of the sample scientific and technological innovation projects; then, the server preprocesses the sample project data to obtain preprocessed sample project data of the sample scientific and technological innovation projects; then, the server performs feature extraction processing on the preprocessed sample project data to obtain a sample feature vector of the preprocessed sample project data, and inputs the sample feature vector into the risk prediction model to be trained to obtain a predicted risk value of the sample scientific and technological innovation project; then, the server obtains the actual risk value of the sample scientific and technological innovation project from the database, and iteratively trains the risk prediction model to be trained according to the difference between the predicted risk value and the actual risk value to obtain a trained risk prediction model.

[0147] In this embodiment, by pre-training the risk prediction model, it is convenient to predict the current risk value of the scientific and technological innovation project in actual application after the project data of the scientific and technological innovation project to be analyzed; moreover, the risk prediction model receives new data in each round of iteration, and performs internal improvements and optimizations on the model, so that predictions can be made more effectively, which is conducive to improving the prediction accuracy of the risk prediction model.

[0148] In an exemplary embodiment, the risk prediction model to be trained is iteratively trained according to the difference between the predicted risk value and the actual risk value to obtain a trained risk prediction model, which specifically includes the following contents: according to the difference between the predicted risk value and the actual risk value, the prediction error value of the risk prediction model to be trained is determined; when the prediction error value is greater than or equal to the error value threshold, the model parameters of the risk prediction model to be trained are adjusted according to the prediction error value to obtain a risk prediction model after the model parameters are adjusted; the risk prediction model after the model parameters are adjusted is trained again until the prediction error value obtained by the trained risk prediction model is less than the error value threshold, then the training is stopped, and the trained risk prediction model is used as the trained risk prediction model.

[0149] The prediction error value is used to indicate the degree of deviation between the predicted risk value and the actual risk value. In actual scenarios, the prediction error value refers to the mean absolute percentage error, which can be calculated using the following formula:

[0150] , formula (1)

[0151] in, is the actual risk value, To predict the risk value, is the sample size.

[0152] The error value threshold refers to a preset error value.

[0153] Among them, model parameters refer to the adjustable internal variables in the risk prediction model.

[0154] Exemplarily, the server determines the difference between the predicted risk value and the actual risk value, determines the prediction error value of the risk prediction model to be trained based on the difference, and judges the prediction error value; when the prediction error value is greater than or equal to the error value threshold, the server adjusts the model parameters of the risk prediction model to be trained based on the prediction error value to obtain the risk prediction model after the model parameters are adjusted; then, the server re-trains the risk prediction model after the model parameters are adjusted until the prediction error value obtained by the trained risk prediction model is less than the error value threshold, then stops training, and uses the trained risk prediction model as the trained risk prediction model.

[0155] In this embodiment, by continuously iterating until the error of the risk prediction model is lower than the threshold, the model's fitting ability on the sample data is ensured, and overfitting is avoided through error threshold control. The final trained model can more accurately capture the risk characteristics of scientific and technological innovation projects, and provide scientific and quantitative support for risk value prediction and subsequent level classification in practical applications.

[0156] In an exemplary embodiment, Figure 2 As shown, another method for handling risks of scientific and technological innovation projects based on artificial intelligence is provided. This method is described by taking the application of the method to a server as an example. Specifically, the method includes the following steps:

[0157] Step S201: Acquire project data of a scientific and technological innovation project to be analyzed; the project data at least includes project background information, project technical indicator data, market data, and project team information of the scientific and technological innovation project.

[0158] Step S202 , preprocessing the project data to obtain preprocessed project data of the scientific and technological innovation project; the preprocessed project data at least includes preprocessed project background information, preprocessed project technical indicator data, preprocessed market data and preprocessed project team information of the scientific and technological innovation project.

[0159] Step S203: determine the data features of the pre-processed project data, and select a target risk prediction model that matches the data features from multiple trained risk prediction models.

[0160] In step S204, feature extraction processing is performed on the preprocessed project background information, the preprocessed project technical indicator data, the preprocessed market data, and the preprocessed project team information, respectively, to obtain a first feature vector of the preprocessed project background information, a second feature vector of the preprocessed project technical indicator data, a third feature vector of the preprocessed market data, and a fourth feature vector of the preprocessed project team information.

[0161] Step S205 , performing fusion processing on the first eigenvector, the second eigenvector, the third eigenvector, and the fourth eigenvector to obtain a eigenvector of the preprocessed project data.

[0162] Step S206: When the current risk value is less than the first set threshold, the current risk level of the scientific and technological innovation project is determined to be a low risk level; when the current risk value is greater than or equal to the first set threshold and less than or equal to the second set threshold, the current risk level of the scientific and technological innovation project is determined to be a medium risk level; when the current risk value is greater than the second set threshold, the current risk level of the scientific and technological innovation project is determined to be a high risk level.

[0163] Step S207: Determine the current risk level of the scientific and technological innovation project based on the current risk value.

[0164] Step S208: Obtain the historical risk level of the scientific and technological innovation project, and determine the future risk level of the scientific and technological innovation project based on the historical risk level and the current risk level.

[0165] Step S209: The current risk level and the future risk level are integrated to obtain the target risk level of the scientific and technological innovation project.

[0166] Step S210: Generate a current risk handling instruction corresponding to the target risk level, and perform corresponding risk handling on the scientific and technological innovation project according to the current risk handling instruction.

[0167] In the above-mentioned artificial intelligence-based risk management method for scientific and technological innovation projects, in the process of risk management of scientific and technological innovation projects, multi-dimensional data collection can ensure the comprehensiveness of the risk assessment basis, improve data quality with the help of data preprocessing, and match the most suitable risk prediction model based on data features to enhance the model's pertinence. Accurate quantification of current risks is achieved through feature extraction and model calculation, and dynamic trend analysis is constructed in combination with historical risk levels. The current future risk level and the future risk level are fused to take into account the immediate status and development situation. Finally, targeted processing instructions are generated based on the fused target risk level, which is conducive to improving the accuracy of risk management of scientific and technological innovation projects. Moreover, the entire process does not require human intervention, avoiding the high subjectivity in the method of using manual experience and simple rules for processing, and the differences in the judgment standards of different evaluators, which leads to the defect of low risk management accuracy of scientific and technological innovation projects, thereby improving the accuracy of risk management of scientific and technological innovation projects.

[0168] In an exemplary embodiment, in order to more clearly illustrate the risk management method for scientific and technological innovation projects based on artificial intelligence provided by the embodiment of the present application, the risk management method for scientific and technological innovation projects based on artificial intelligence is specifically described below with a specific embodiment. In one embodiment, Figure 3 As shown, this application also provides a risk prediction method for scientific and technological innovation projects based on artificial intelligence. Specifically, it includes the following contents:

[0169] Obtain data related to scientific and technological innovation projects, including multi-dimensional data such as project background, technical indicators, market data, and team information, to construct the original data set;

[0170] Use data preprocessing technology to clean, remove duplicates, and standardize raw data to improve data quality;

[0171] Build a risk prediction model based on artificial intelligence algorithms, input preprocessed data into the model for training, and determine the key features and prediction rules for risk prediction;

[0172] Use the trained risk prediction model to predict the risks of scientific and technological innovation projects and output the risk level assessment results;

[0173] By comparing the prediction results with the actual project risk situation, feedback is provided to optimize the risk prediction model;

[0174] Among them, when constructing a risk prediction model, the artificial intelligence algorithms that can be used include but are not limited to logistic regression algorithm, support vector machine algorithm, and random forest algorithm.

[0175] Specifically, this method first collects data related to scientific and technological innovation projects from multiple dimensions, including project background, technical indicators, market data, and team information, to construct an original data set. This provides a basic data source for subsequent analysis. Then, data preprocessing technology is used to clean, deduplicate, and standardize the original data to improve data quality and ensure that the data input into the model is accurate and reliable. A risk prediction model is then constructed based on an artificial intelligence algorithm, and the preprocessed data is input into the model training to determine the key features and rules for risk prediction. The trained model is used to predict project risks and output risk level assessment results. Finally, by comparing the predicted results with the actual project risk situation, feedback is provided to optimize the risk prediction model. When constructing the model, a variety of artificial intelligence algorithms can be used, such as logistic regression algorithm, support vector machine algorithm, random forest algorithm, etc.

[0176] Multi-dimensional data collection can comprehensively reflect project status and avoid risk misjudgments due to missing data. Data preprocessing improves data quality and lays the foundation for accurate model training. Building models based on artificial intelligence algorithms leverages the advantages of these algorithms to uncover potential relationships in the data and improve the accuracy of risk predictions. Outputting risk level assessment results provides project decision makers with intuitive risk information, facilitating decision-making. Feedback optimization mechanisms enable the model to continuously adapt to new data and project changes, continuously improving predictive performance.

[0177] Obtaining relevant data on scientific and technological innovation projects and constructing the original data set specifically includes:

[0178] Collect project background information through multiple channels, including the original intention of the project and the direction of technical research and development;

[0179] Collect project technical indicator data, including technical innovation indicators and technical maturity indicators;

[0180] Collect market data, including target market size, market growth rate, and competitor situation;

[0181] Obtain project team information, including team members' professional background and project experience, and integrate this data into the original data set.

[0182] Specifically, multi-channel collection ensures comprehensive information, covering all key aspects of the project. Targeted data collection accurately reflects the true status of the project in terms of technology, market, team, and other dimensions, providing a rich and accurate data foundation for subsequent risk prediction, thereby improving the reliability and comprehensiveness of risk prediction.

[0183] The technical indicator data of the collected projects specifically include:

[0184] For technological innovation indicators, quantitative data is collected by analyzing the degree of technological leadership;

[0185] For technology maturity indicators, data is collected from the aspects of technology feasibility verification and technology stability assessment.

[0186] Specifically, the technical innovation index reflects, to a certain extent, the high level of technological innovation activity of the project. The degree of technological leadership is determined by comparing and analyzing similar technologies in the industry, such as technical performance indicators and application scenario expansion, to determine the leading position of the project technology in the industry. For the technical maturity index, the degree of technical feasibility verification determines the feasibility of the technology from theoretical to practical application by considering the success rate of technical principle verification experiments and the operational performance of the technology in simulated actual scenarios. The technical stability assessment evaluates the stability of the technology in actual operation by considering the failure rate of the technical system over a certain period of time and performance fluctuations.

[0187] Quantitatively collecting technical indicator data can transform abstract technical characteristics into measurable data, facilitating accurate assessment of project technical risks. A comprehensive assessment of technological innovation and maturity helps identify potential technical risks. For example, insufficient technological innovation may lead to market competition, while low technological maturity may lead to technical difficulties during project implementation, thus providing key technical dimension information for project risk prediction.

[0188] Use data preprocessing technology to clean, remove duplicates, and standardize the original data, including:

[0189] Use data cleaning algorithms to identify and delete duplicate data, erroneous data, and data records with excessive missing values ​​in the original data;

[0190] Use hash algorithm to remove duplicates and ensure the uniqueness of data;

[0191] The numerical data are standardized using the Z-Score standardization formula.

[0192] Specifically, cleaning improves data accuracy, deduplication reduces redundancy, and standardization makes data comparable, providing high-quality data for model training and improving training efficiency and prediction accuracy.

[0193] Building a risk prediction model based on artificial intelligence algorithms and inputting preprocessed data into the model for training specifically includes:

[0194] If the data features are linearly separable, the logistic regression algorithm is preferred to construct the risk prediction model, and the maximum likelihood estimation is used as the objective function for model training;

[0195] If the data has complex nonlinear boundary characteristics, the support vector machine algorithm is used to map low-dimensional data to high-dimensional space through kernel functions to build and train the model;

[0196] When the data feature dimension is high and there is a lot of noise, the random forest algorithm is used to build a risk prediction model by training multiple decision trees in parallel.

[0197] Specifically, flexible algorithm selection based on data characteristics can fully leverage the strengths of different algorithms. Logistic regression algorithms are computationally efficient when processing linearly separable data and can clearly identify key features and rules for risk prediction. Support vector machine algorithms effectively handle complex nonlinear data relationships, improving the model's adaptability to complex data. The random forest algorithm, through the parallel operation of multiple decision trees, enhances the model's noise tolerance and stability, improving the accuracy and reliability of risk predictions and adapting to the risk prediction needs of different data types.

[0198] The training of logistic regression model with maximum likelihood estimation as the objective function specifically includes:

[0199] Assume that the training data set is ,in For the The feature vector of the samples, For sample labels, the likelihood function of the logistic regression model is:

[0200] , formula (2)

[0201] The gradient descent method is used to solve the problem The largest and .

[0202] Specifically, maximum likelihood estimation starts from a probability perspective, allowing the model to fit the true distribution of the data as much as possible during the training process. By solving it through optimization algorithms such as gradient descent, it can efficiently find the optimal parameters, improve the risk prediction accuracy of the logistic regression model for linearly separable data, and provide reliable linear model support for project risk prediction.

[0203] Use the trained risk prediction model to predict the risks of scientific and technological innovation projects and output the risk level assessment results, including:

[0204] Set risk level classification standards, including low risk, medium risk, and high risk;

[0205] The data of the project to be predicted is input into the trained risk prediction model, and the model outputs the predicted risk value;

[0206] The risk level of the project is determined and output by matching the risk value with the set risk level classification standards.

[0207] Specifically, we first set the risk classification criteria, which are divided into three levels: low risk, medium risk, and high risk. The data of the project to be predicted is input into the trained risk prediction model. The model calculates and analyzes the input data based on the key features and prediction rules obtained through training, and outputs a predicted risk value. This risk value is then matched with the set risk classification criteria to determine the level within which the risk value falls. This determines the project's risk level and outputs the predicted risk value.

[0208] Clear risk grading standards provide a clear metric for project risk assessment. The model accurately outputs risk values ​​and matches them to grading levels, providing project decision-makers with intuitive and easy-to-understand risk information. This allows them to quickly understand project risk status, facilitate the development of targeted risk response strategies, and improve the efficiency and scientific nature of project decision-making.

[0209] The risk classification standards include:

[0210] When the risk value output by the risk prediction model is less than the set threshold T1, the project is judged to be at a low risk level;

[0211] When the risk value is between T1 and T2, the project is judged to be of medium risk level;

[0212] When the risk value is greater than T2, the project is judged to be at a high risk level, where the values ​​of T1 and T2 are determined through statistical analysis of a large amount of historical project data.

[0213] Specifically, a large amount of historical project data is statistically analyzed to determine the correspondence between the output value of the risk prediction model and the actual risk status, and T1 and T2 thresholds are set to divide the risk levels. The standards determined based on historical data are scientific and objective, and the threshold setting provides a basis for risk judgment, avoids subjective arbitrariness, and helps to accurately assess risks.

[0214] By comparing the prediction results with the actual project risk situation, the feedback optimization risk prediction model specifically includes:

[0215] Calculate the error between the predicted risk value and the actual risk value using the mean absolute percentage error:

[0216] , formula (1)

[0217] in, is the actual risk value, To predict the risk value, is the sample size;

[0218] Determine the deviation of the risk prediction model based on the size of the error;

[0219] If the error exceeds the set acceptable range, adjust the parameters of the risk prediction model;

[0220] Reuse historical data to train the adjusted model until the model prediction error meets the requirements.

[0221] Specifically, the mean absolute percentage error (MAPE) provides a direct reflection of the degree of deviation between the predicted and actual values. By adjusting model parameters through error feedback, the model can be continuously optimized, improving forecast accuracy. Continuous training using historical data allows the model to better adapt to the data characteristics of different projects, enhancing its generalization capabilities and enabling it to more accurately predict the risks of new scientific and technological innovation projects.

[0222] After adjusting the parameters of the risk prediction model, reusing historical data to train the model specifically includes:

[0223] Randomly extract a certain percentage of data from the historical project dataset as the training set, and the remaining data as the test set;

[0224] Input the training set data into the risk prediction model with adjusted parameters for training;

[0225] Use the test set data to evaluate the performance of the trained model. If the prediction error of the model on the test set still does not meet the requirements, further adjust the parameters and repeat the training and evaluation process.

[0226] Specifically, randomly dividing the training and test sets ensures the randomness and representativeness of the data, making model training and evaluation more objective. Through continuous training and evaluation, model parameters are gradually optimized, improving model performance, making it better able to cope with data from different scientific and technological innovation projects, and improving the accuracy and stability of risk prediction, providing more reliable model support for project risk prediction.

[0227] In the above embodiment, in the process of risk management of scientific and technological innovation projects, the comprehensiveness of the risk assessment basis can be ensured through multi-dimensional data collection, the data quality can be improved with the help of data preprocessing, and the most suitable risk prediction model can be matched based on data features to enhance the model's pertinence. The current risk can be accurately quantified through feature extraction and model calculation, and dynamic trend analysis can be constructed in combination with historical risk levels. The current future risk level and the future risk level are fused to take into account the immediate status and development situation. Finally, targeted processing instructions are generated based on the fused target risk level, which is conducive to improving the accuracy of risk management of scientific and technological innovation projects. Moreover, the entire process does not require human intervention, avoiding the high subjectivity in the method of using manual experience and simple rules for processing, and the differences in the judgment standards of different evaluators, which leads to the defect of low risk management accuracy of scientific and technological innovation projects, thereby improving the risk management accuracy of scientific and technological innovation projects.

[0228] It should be understood that, although the steps in the flowcharts of the above embodiments are shown in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the flowcharts of the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily to be performed in sequence, but can be performed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0229] Based on the same inventive concept, the embodiments of the present application also provide an AI-based scientific and technological innovation project risk management device for implementing the aforementioned AI-based scientific and technological innovation project risk management method. The implementation solution provided by this device is similar to the implementation solution described in the aforementioned method. Therefore, the specific limitations of one or more AI-based scientific and technological innovation project risk management device embodiments provided below can be found in the limitations of the AI-based scientific and technological innovation project risk management method above, and will not be repeated here.

[0230] In an exemplary embodiment, Figure 4 As shown, a risk processing device for scientific and technological innovation projects based on artificial intelligence is provided, comprising: a data acquisition module 401, a data processing module 402, a model screening module 403, a risk prediction module 404, a first determination module 405, a second determination module 406, a level fusion module 407 and a risk processing module 408, wherein:

[0231] The data acquisition module 401 is used to acquire the project data of the scientific and technological innovation project to be analyzed; the project data at least includes the project background information, project technical indicator data, market data and project team information of the scientific and technological innovation project.

[0232] The data processing module 402 is used to pre-process the project data to obtain pre-processed project data of the scientific and technological innovation project.

[0233] The model screening module 403 is used to determine the data features of the pre-processed project data and screen out a target risk prediction model that matches the data features from multiple trained risk prediction models.

[0234] The risk prediction module 404 is used to perform feature extraction processing on the pre-processed project data to obtain a feature vector of the pre-processed project data, and input the feature vector into the target risk prediction model to obtain the current risk value of the scientific and technological innovation project.

[0235] The first determination module 405 is used to determine the current risk level of the scientific and technological innovation project according to the current risk value.

[0236] The second determination module 406 is used to obtain the historical risk level of the scientific and technological innovation project, and determine the future risk level of the scientific and technological innovation project based on the historical risk level and the current risk level.

[0237] The level fusion module 407 is used to fuse the current risk level and the future risk level to obtain the target risk level of the scientific and technological innovation project.

[0238] The risk processing module 408 is used to generate a current risk processing instruction corresponding to the target risk level, and perform corresponding risk processing on the scientific and technological innovation project according to the current risk processing instruction.

[0239] In an exemplary embodiment, the risk prediction module 404 is also used to perform feature extraction processing on the preprocessed project background information, the preprocessed project technical indicator data, the preprocessed market data and the preprocessed project team information, respectively, to obtain a first feature vector of the preprocessed project background information, a second feature vector of the preprocessed project technical indicator data, a third feature vector of the preprocessed market data, and a fourth feature vector of the preprocessed project team information; and to perform fusion processing on the first feature vector, the second feature vector, the third feature vector and the fourth feature vector to obtain a feature vector of the preprocessed project data.

[0240] In an exemplary embodiment, the data processing module 402 is also used to determine the hash value of each sub-project data in the project data, and determine the duplicate data in the project data based on the hash value of each sub-project data; delete the duplicate data in the project data to obtain deduplicated project data; normalize the deduplicated project data to obtain normalized project data as preprocessed project data for the scientific and technological innovation project.

[0241] In an exemplary embodiment, the first determination module 405 is also used to determine that the current risk level of the scientific and technological innovation project is a low risk level when the current risk value is less than the first set threshold; to determine that the current risk level of the scientific and technological innovation project is a medium risk level when the current risk value is greater than or equal to the first set threshold and less than or equal to the second set threshold; and to determine that the current risk level of the scientific and technological innovation project is a high risk level when the current risk value is greater than the second set threshold.

[0242] In an exemplary embodiment, the artificial intelligence-based scientific and technological innovation project risk processing device also includes a model training module, which is used to obtain sample project data of sample scientific and technological innovation projects; preprocess the sample project data to obtain preprocessed sample project data of the sample scientific and technological innovation projects; perform feature extraction processing on the preprocessed sample project data to obtain a sample feature vector of the preprocessed sample project data, and input the sample feature vector into the risk prediction model to be trained to obtain a predicted risk value of the sample scientific and technological innovation project; obtain the actual risk value of the sample scientific and technological innovation project, and iteratively train the risk prediction model to be trained based on the difference between the predicted risk value and the actual risk value to obtain a trained risk prediction model.

[0243] In an exemplary embodiment, the model training module is also used to determine the prediction error value of the risk prediction model to be trained based on the difference between the predicted risk value and the actual risk value; when the prediction error value is greater than or equal to the error value threshold, the model parameters of the risk prediction model to be trained are adjusted according to the prediction error value to obtain a risk prediction model after the model parameters are adjusted; the risk prediction model after the model parameters are adjusted is trained again until the prediction error value obtained by the trained risk prediction model is less than the error value threshold, then the training is stopped, and the trained risk prediction model is used as the trained risk prediction model.

[0244] Each module in the aforementioned AI-based scientific and technological innovation project risk management device may be implemented in whole or in part through software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor in a computer device in hardware form, or may be stored in a computer device's memory in software form, so that the processor can call and execute the corresponding operations of each module.

[0245] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 5As shown. The computer device includes a processor, a memory, an input / output interface (I / O) and a communication interface. The processor, memory and input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store project background information, project technical indicator data, market data and project team information, etc. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a risk management method for scientific and technological innovation projects based on artificial intelligence.

[0246] Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0247] In an exemplary embodiment, a computer device is further provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0248] In an exemplary embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0249] In an exemplary embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0250] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments. In particular, any reference to memory, database, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processors involved in the various embodiments provided herein may be, but are not limited to, general-purpose processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), programmable logic devices (PLDs), data processing logic devices based on quantum computing, and the like.

[0251] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0252] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A method for handling risks of scientific and technological innovation projects based on artificial intelligence, characterized in that: The method comprises: Obtaining project data of the scientific and technological innovation project to be analyzed; the project data at least includes project background information, project technical indicator data, market data, and project team information of the scientific and technological innovation project; Preprocessing the project data to obtain preprocessed project data of the scientific and technological innovation project; Determining data features of the preprocessed project data, and screening a target risk prediction model that matches the data features from a plurality of trained risk prediction models; Performing feature extraction processing on the pre-processed project data to obtain a feature vector of the pre-processed project data, and inputting the feature vector into the target risk prediction model to obtain a current risk value of the scientific and technological innovation project; Determining the current risk level of the scientific and technological innovation project based on the current risk value; Obtaining a historical risk level of the scientific and technological innovation project, and determining a future risk level of the scientific and technological innovation project based on the historical risk level and the current risk level; Fusing the current risk level and the future risk level to obtain a target risk level for the scientific and technological innovation project; Generate a current risk handling instruction corresponding to the target risk level, and perform corresponding risk handling on the scientific and technological innovation project according to the current risk handling instruction.

2. The method according to claim 1, characterized in that The pre-processed project data at least includes the pre-processed project background information, pre-processed project technical indicator data, pre-processed market data and pre-processed project team information of the scientific and technological innovation project; The performing feature extraction processing on the preprocessed project data to obtain a feature vector of the preprocessed project data includes: performing feature extraction processing on the preprocessed project background information, the preprocessed project technical indicator data, the preprocessed market data, and the preprocessed project team information, respectively, to obtain a first feature vector of the preprocessed project background information, a second feature vector of the preprocessed project technical indicator data, a third feature vector of the preprocessed market data, and a fourth feature vector of the preprocessed project team information; The first eigenvector, the second eigenvector, the third eigenvector and the fourth eigenvector are fused to obtain a eigenvector of the preprocessed project data.

3. The method according to claim 1, characterized in that The preprocessing of the project data to obtain the preprocessed project data of the scientific and technological innovation project includes: Determining a hash value of each sub-item data in the item data, and determining duplicate data in the item data based on the hash value of each sub-item data; Deleting duplicate data from the project data to obtain deduplicated project data; The deduplicated project data is normalized to obtain normalized project data, which serves as preprocessed project data for the scientific and technological innovation project.

4. The method according to claim 1, wherein Determining the current risk level of the scientific and technological innovation project based on the current risk value includes: When the current risk value is less than a first set threshold, determining that the current risk level of the scientific and technological innovation project is a low risk level; When the current risk value is greater than or equal to the first set threshold and less than or equal to the second set threshold, determining that the current risk level of the scientific and technological innovation project is a medium risk level; When the current risk value is greater than the second set threshold, it is determined that the current risk level of the scientific and technological innovation project is a high risk level.

5. The method according to any one of claims 1 to 4, characterized in that Each trained risk prediction model is trained in the following way: Obtain sample project data of sample scientific and technological innovation projects; Preprocessing the sample project data to obtain preprocessed sample project data of the sample scientific and technological innovation project; Performing feature extraction processing on the preprocessed sample project data to obtain a sample feature vector of the preprocessed sample project data, and inputting the sample feature vector into a risk prediction model to be trained to obtain a predicted risk value of the sample scientific and technological innovation project; The actual risk value of the sample scientific and technological innovation project is obtained, and according to the difference between the predicted risk value and the actual risk value, the risk prediction model to be trained is iteratively trained to obtain the trained risk prediction model.

6. The method according to claim 5, characterized in that The iterative training of the risk prediction model to be trained according to the difference between the predicted risk value and the actual risk value to obtain the trained risk prediction model includes: Determining a prediction error value of the risk prediction model to be trained based on a difference between the predicted risk value and the actual risk value; When the prediction error value is greater than or equal to the error value threshold, adjusting the model parameters of the risk prediction model to be trained according to the prediction error value to obtain a risk prediction model after the model parameters are adjusted; The risk prediction model after the model parameters are adjusted is trained again until the prediction error value obtained by the trained risk prediction model is less than the error value threshold, then the training is stopped, and the trained risk prediction model is used as the trained risk prediction model.

7. A risk management device for scientific and technological innovation projects based on artificial intelligence, characterized in that: The device comprises: A data acquisition module is used to acquire project data of the scientific and technological innovation project to be analyzed; the project data at least includes project background information, project technical indicator data, market data and project team information of the scientific and technological innovation project; A data processing module, configured to pre-process the project data to obtain pre-processed project data of the scientific and technological innovation project; A model screening module is used to determine the data characteristics of the pre-processed project data and screen out a target risk prediction model that matches the data characteristics from multiple trained risk prediction models; a risk prediction module, configured to perform feature extraction processing on the pre-processed project data to obtain a feature vector of the pre-processed project data, and input the feature vector into the target risk prediction model to obtain a current risk value of the scientific and technological innovation project; A first determination module is configured to determine a current risk level of the scientific and technological innovation project based on the current risk value; A second determination module is configured to obtain a historical risk level of the scientific and technological innovation project and determine a future risk level of the scientific and technological innovation project based on the historical risk level and the current risk level; A level fusion module, configured to fuse the current risk level and the future risk level to obtain a target risk level for the scientific and technological innovation project; The risk processing module is used to generate a current risk processing instruction corresponding to the target risk level, and perform corresponding risk processing on the scientific and technological innovation project according to the current risk processing instruction.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images