New material price trend analysis method and system based on big data

Abstract

The invention discloses a new material price trend analysis method and system based on big data, and relates to the technical field of price trend analysis, and the processing method comprises the specific steps: market data monitoring: monitoring and collecting market data in real time, carrying out the continuous evaluation and optimization of a data source, guaranteeing the quality and diversity of the data, and carrying out the real-time monitoring of the market data; the data fusion technology is used for integrating information from different sources, the reliability of the data is improved, the market data is monitored and collected in real time, the data fusion technology is used for integrating the information from the different sources, the reliability and diversity of the data are greatly improved, the accuracy and comprehensiveness of price trend analysis of new materials are ensured, and the market competitiveness is improved. According to the method, more reliable data support is provided for decision makers, the accuracy and integrity of data are ensured through regular data quality monitoring and evaluation, and the reliability of analysis results is further enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of price trend analysis, and in particular to a new material price trend analysis method and system based on big data. Background Art

[0002] Traditional market analysis methods have certain reference value, but in the face of an increasingly complex and changing market environment and massive amounts of data, traditional methods have gradually revealed their limitations. Traditional price trend analysis methods often rely on limited market data and empirical judgments, lack comprehensiveness and accuracy, and result in large deviations in analysis results.

[0003] Secondly, with the rapid development of new material technology, the relationship between material composition, performance and cost is becoming increasingly complex. Traditional analysis methods are difficult to effectively capture these subtle and critical changes. In analyzing the impact of material composition on price, traditional methods often only consider a few factors and ignore other potential important influencing factors, resulting in incomplete and one-sided analysis results.

[0004] In addition, traditional price trend analysis methods lack dynamism and real-time performance. Market environment and consumer demand change rapidly and frequently, and price trends can only be reviewed and summarized after the fact. Effective decision-making support cannot be provided in a timely manner, resulting in companies being slow to respond to market changes and unable to make timely adjustments, thus missing out on business opportunities.

[0005] Therefore, a new material price trend analysis method based on big data is developed to solve the above problems. Summary of the invention

[0006] The purpose of the present invention is to make up for the shortcomings of the existing technology and provide a new material price trend analysis method and system based on big data, which can accurately predict the price trend of new materials by real-time monitoring of market data, evaluating the influence of material composition, setting heat coefficient and integrating multivariate data to establish a price prediction model. It not only improves the data quality and diversity, but also provides decision support reports for decision makers through intelligent analysis, reduces procurement and investment risks, has important application value in the field of material research and development and market analysis, and provides strong technical support for the research and development and market application of new materials.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions: On the one hand, a new material price trend analysis method based on big data is characterized by the following specific steps:

[0008] S100, Market Data Monitoring: Real-time monitoring and collection of market data, continuous evaluation and optimization of data sources to ensure data quality and diversity, and use data fusion technology to integrate information from different sources to improve data reliability;

[0009] S200, Material composition impact assessment: Collect composition information from multiple aspects, create a database to store composition data, analyze the impact of composition on material performance and cost, and identify the components that have a significant impact on price;

[0010] S300, heat coefficient setting: determine the indicators that reflect the market's attention to new materials, collect market heat data through multiple channels, calculate the heat coefficient of each indicator based on the collected data, comprehensively obtain the overall heat coefficient, and establish a dynamic adjustment mechanism for the heat coefficient to adapt to changes in the market environment;

[0011] S400, price trend forecast: Integrate raw material prices, heat coefficients and market trend data, study and analyze the collected data, establish a price forecast model, verify the forecast results through historical data, and analyze price trends and potential market influencing factors;

[0012] S500, Decision Support Report Generation: Generate reports based on analysis results, evaluate the uncertainty and potential risks of forecast results, provide procurement and investment recommendations based on analysis results, and present the reports to decision makers in an appropriate format.

[0013] Further, the S100 includes:

[0014] S101, identify key market data indicators and evaluate the reliability, coverage and update frequency of data sources;

[0015] S102, using web crawler technology to capture data in real time to ensure the timeliness and continuity of data collection;

[0016] S103, clean the collected data, remove duplications, correct errors, standardize formats, apply data fusion technology, integrate data from different sources, and enhance data consistency;

[0017] S104, storing the cleaned data in a database and establishing an efficient data index and query mechanism;

[0018] S105, regularly monitor and evaluate data quality to ensure data accuracy and completeness.

[0019] Furthermore, the S200 includes:

[0020] S201, collect relevant data on ingredients from scientific literature, laboratory reports, and material specifications;

[0021] S202, creating a relational database based on the ingredient table, performance table, and cost table to store the ingredient data;

[0022] S203, preprocessing the component data to remove duplicate records, process missing values, and standardize, and merge component data from different sources;

[0023] S204, using the Pearson correlation coefficient to explore the relationship between composition and performance, and to analyze the impact of changes in composition ratios on material production costs;

[0024] S205, applying a decision tree to identify key factors affecting the price.

[0025] Furthermore, in S204, the relationship between the components and the performance is explored using the Pearson correlation coefficient. Based on the collected component data and performance data, the Pearson correlation coefficient between the components and the performance is calculated. The mathematical expression of the Pearson correlation coefficient is: Where: X i and Y i are the observed values ​​of composition and performance, and is their mean, n is the sample size, r xy The absolute value of close to 1 indicates a strong linear relationship, and close to 0 indicates a weak linear relationship. In order to determine whether the observed correlation is significant, a significance test is performed. The calculation formula of the statistic t of the correlation coefficient is: Among them: r is the correlation coefficient, n is the sample size, and when the value of the statistic is greater than the significance level α, it is considered that there is a significant correlation between the two variables.

[0026] Furthermore, in S205, a decision tree is used to identify key components that affect prices, and a data set containing new material prices, material components, and other features that affect prices is collected. The collected data set is cleaned to remove duplicate data, correct erroneous data, and process missing values ​​and outliers. The information gain of each feature is calculated to select features related to prices from the data set. The information gain formula is: Where: D is the data set, A is the attribute, Val(A) is the set of values ​​of attribute A, D v It is to select the data subset after the attribute A value is υ, select the optimal feature for splitting, and divide the data set into different subsets based on the selected optimal feature. Repeat the above feature selection and splitting process for each subset until the stopping condition is met, and a complete decision tree model is constructed. In order to prevent overfitting, it is necessary to prune the constructed decision tree, collect real-time data to evaluate the performance of the decision tree model, analyze the decision tree results, and identify the key components that affect the material price.

[0027] Furthermore, the S300 includes:

[0028] S301, determine the indicators that reflect the market's attention to new materials;

[0029] S302, collects data from multiple channels including search engine APIs, social media platforms, news websites, industry reports, online markets, and forums;

[0030] S303, assigning corresponding weights according to the influence of each indicator on market attention, and calculating weighted heat coefficient according to the weights;

[0031] S304, combining the heat coefficients of the various indicators to obtain an overall heat coefficient;

[0032] S305, design a dynamic adjustment mechanism to update the heat coefficient based on market feedback and new data;

[0033] S306, use historical data to verify the effectiveness of the heat coefficient model and make adjustments based on market feedback.

[0034] Furthermore, the S400 includes:

[0035] S401, collects data on raw material prices, heat coefficients, and market trends, integrates data from different sources and types, and builds a comprehensive data set;

[0036] S402, cleaning the data, processing missing values, outliers and duplicate records, and performing standardization transformation on the cleaned data to eliminate the influence of different dimensions;

[0037] S403, applying feature selection methods to screen features that have a significant impact on prediction, and dividing the data into a training set and a validation set;

[0038] S404, selecting a linear regression model to predict future price trends based on data characteristics and prediction targets;

[0039] S405 uses the training set to train the selected model, adjusts the model parameters, improves the prediction accuracy, and uses the validation set to evaluate the stability and generalization ability of the model using the cross-validation technique;

[0040] S406, using the trained model to perform price prediction on the latest market data, and generating prediction results of future price trends based on the model output.

[0041] Furthermore, the S404 predicts future price trends by analyzing historical data using a linear regression model, extracting data with significant features into a training set and a validation set, analyzing the distribution characteristics of the data, evaluating the time span and time resolution of the data, determining the predicted time range, and clarifying the application scenarios of the prediction results. The formula of the linear regression model is: y=β0+β1x1+β2x2+...+β n x n +∈, where: y is the predicted value, x is the raw material price and heat coefficient, β0 is the intercept term, β n is the coefficient, ∈ is the error term, and the model parameters are adjusted by minimizing the loss function using the training set. The mathematical formula for minimizing the sum of squares of the error terms is: Where: m is the number of samples, y is i is the actual value, is the predicted value. The training set is used to adjust the model parameters by minimizing the loss function. The validation set is used to evaluate the prediction performance of the model. The model is adjusted according to the evaluation results, and the trained model is used to predict future prices.

[0042] Furthermore, the S500 includes:

[0043] S501, summarizing the results of the steps of price trend forecasting and market influencing factor analysis;

[0044] S502, using machine learning models to assess the uncertainty of prediction results;

[0045] S503, identify and assess potential market risks, supply risks, and technical risks;

[0046] S504, formulate procurement and investment recommendations based on analysis results and risk assessment;

[0047] S505, integrate analysis results, uncertainty assessment, risk analysis and recommendations into a report.

[0048] On the other hand, a new material price trend analysis system based on big data is characterized in that the system includes:

[0049] Data collection module: responsible for collecting multi-dimensional data such as new material market data, R&D data, and production data, and integrating the data into a unified format and storing it in the database;

[0050] Data preprocessing module: Data preprocessing module: performs a series of preprocessing operations on the collected data, including cleaning, deduplication, and missing value filling, to ensure the quality and accuracy of the data;

[0051] Feature extraction module: extracts key features of material type, production process, market demand, and raw material price from preprocessed data as the basis for subsequent prediction and analysis;

[0052] Prediction model building module: Use deep learning algorithms to create new material price trend prediction models, use historical data and extracted features for training and optimization to improve prediction accuracy;

[0053] Result display module: The prediction results are displayed to users in a visual form. Users can adjust R&D strategies and market strategies based on the prediction results to improve R&D efficiency and engineering application level.

[0054] Compared with the existing technology, this new material price trend analysis method based on big data has the following advantages:

[0055] Beneficial effects:

[0056] 1. The present invention monitors and collects market data in real time and uses data fusion technology to integrate information from different sources, which greatly improves the reliability and diversity of data, ensures the accuracy and comprehensiveness of new material price trend analysis, and provides more reliable data support for decision makers. Through regular data quality monitoring and evaluation, the accuracy and completeness of the data are guaranteed, and the reliability of the analysis results is further enhanced.

[0057] 2. The present invention adopts an intelligent prediction model to predict and analyze the price of new materials by integrating raw material prices, heat coefficients and market trend data. This not only improves the R&D efficiency and engineering application level, but also provides a scientific basis for procurement and investment recommendations. In particular, in terms of the evaluation of the impact of material composition, the Pearson correlation coefficient and decision tree algorithm are used to deeply analyze the impact of composition on material performance and cost, identify the key components that affect the price, and provide strong technical support for the R&D and market application of new materials.

[0058] Other advantages, objectives and features of the present invention will be set forth in part in the following description and, in part, will be apparent to those skilled in the art based on an examination of the following or may be taught from the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0060] Figure 1The figure is a flow chart of a new material price trend analysis method based on big data. DETAILED DESCRIPTION

[0061] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] Example 1

[0063] A new material price trend analysis method based on big data, the method comprising the following steps: market data monitoring, material composition impact assessment, heat coefficient setting, price trend forecasting, and decision support report generation.

[0064] First, enter the market data monitoring stage (S100), analyze the key market data indicators required for the price trend of new materials, use web crawler technology, write specific crawling scripts, monitor and capture data of selected data sources in real time, clean the captured raw data, remove duplicate data, correct erroneous data, and unify the data format. The cleaned and fused data are stored in a high-performance database, and an efficient data indexing and query mechanism is established. The data quality is monitored and evaluated regularly to ensure the accuracy and completeness of the data.

[0065] Next, enter the material composition impact assessment stage (S200), collect the composition type, proportion, source and historical price data of the new material from scientific literature, experimental reports, and material specifications, create a relational database to store these data based on the collected composition table, performance table, and cost table, pre-process the stored composition data, including removing duplicate records, processing missing values, standardizing data, and merging composition data from different sources to ensure data accuracy and consistency, and use the Pearson correlation coefficient to explore the relationship between composition and material performance. Based on the collected composition data and performance data, calculate the Pearson correlation coefficient between each component and performance. The mathematical expression of the Pearson correlation coefficient is: By calculating the statistic and comparing it with the significance level α, we can determine whether the observed correlation is significant. This will help us understand which components have an important impact on material properties. We collect data sets containing new material prices, material components, and other features that affect prices. We clean the collected data sets, remove duplicate data, correct erroneous data, and handle missing values ​​and outliers to ensure the accuracy and reliability of the data. By calculating the information gain of each feature, the information gain Select the features most relevant to price from the data set. Based on the selected optimal features, use the decision tree algorithm to split the data set into different subsets. Repeat this process until the stopping condition is met to build a complete decision tree model. In order to prevent overfitting, the constructed decision tree is pruned. The validation set is used to evaluate the performance of the decision tree model, and the model parameters are adjusted according to the evaluation results. The results of the decision tree are analyzed to identify the key components that affect the price of new materials.

[0066] Subsequently, the heat coefficient setting stage (S300) is entered to determine a series of indicators that can reflect the market's attention to new materials. Relevant data of the above indicators are collected from multiple channels through search engine APIs, social media platform APIs, news website RSS subscriptions or APIs, official websites of industry reports, online markets, APIs and forum crawlers. According to the degree of influence of each indicator on market attention, corresponding weights are assigned, and the weighted heat coefficients of all indicators are combined to obtain the overall heat coefficient. In order to maintain the timeliness and accuracy of the heat coefficient model, it is necessary to design a dynamic adjustment mechanism and use historical data to verify the effectiveness of the heat coefficient model.

[0067] Then, the price trend prediction stage (S400) is entered. Data on raw material prices, heat coefficients, and market trends are collected from multiple sources and types to build a comprehensive data set. The collected data is cleaned, missing values, outliers, and duplicate records are processed, and feature selection methods are applied to screen out features that have a significant impact on the prediction from the integrated data set. The feature selection method is applied to screen out features that have a significant impact on the prediction from the integrated data set. According to the data characteristics and prediction goals, a linear regression model is selected to predict future price trends. The extracted data with significant features are divided into a training set and a test set. The distribution characteristics of the data are analyzed, the time span and time resolution of the data are evaluated, the time range of the prediction is determined, and the application scenario of the prediction result is clarified. The formula of the linear regression model is: y=β0+β1x1+β2x2+...+β n x n +∈, use the training set to adjust the model parameters by minimizing the loss function, and the formula for minimizing the sum of squares of the error terms is: The training set is used to adjust the model parameters by minimizing the loss function (MSE), and the test set is used to evaluate the model's prediction performance. The model parameters are adjusted based on the evaluation results, and the cross-validation technique is used to evaluate the model's stability and generalization ability. The model parameters with the best performance are selected by repeatedly training and testing the model, and the trained linear regression model is used to predict the price of the latest market data.

[0068] Finally, enter the decision support report generation stage (S500), summarize the results of price trend forecasting and market influencing factor analysis steps, use statistical methods to evaluate the uncertainty of price trend forecast results, identify and evaluate potential market risks, supply risks and technical risks based on the summarized analysis results, and formulate recommendations for new material procurement and investment based on the analysis results and risk assessment, integrating the analysis results, uncertainty assessment, risk analysis and recommendations into a complete report.

[0069] Example 2

[0070] The article focuses on the application of price trend forecasting in a new material price trend analysis method based on big data:

[0071] Collect data on raw material prices, popularity factors (such as social media discussion volume, search engine search volume), and market trends from multiple channels (such as suppliers, market research companies, and government agencies), unify the types and standardize the formats of the collected data to ensure data consistency and comparability, and integrate the processed data into a comprehensive data set for subsequent analysis.

[0072] Identify and process missing values, outliers, and duplicate records in the data. For missing values, interpolation and mean replacement methods can be used to fill them. For outliers, they are identified and processed according to business logic. For duplicate records, they are directly deleted or merged. Since the dimensions of the data of raw material prices and heat coefficients are different, Z-score standardization is used to perform standardized transformation to eliminate the impact of dimensions.

[0073] Perform feature analysis on the cleaned and standardized data to understand the relationship between each feature and the target variable (new material price). Use statistical methods to perform feature selection and screen out features that have a significant impact on the prediction.

[0074] According to the data characteristics and prediction objectives, a linear regression model is selected for price trend prediction. A linear regression model is constructed based on the selected features. The data set is divided into a training set and a test set. The training set is used to train the model and adjust the model parameters. The test set is used to evaluate the prediction performance of the model. The model parameters are adjusted by minimizing the loss function using the training set.

[0075] Input the latest collected data on raw material prices and heat coefficients into the trained model to obtain prediction results, generate a forecast report or chart of future price trends, and present the prediction results in a visual form to relevant decision makers or users.

[0076] In summary, this method can achieve accurate prediction of new material prices and provide valuable decision-making support for related companies and individuals. With the continuous development of big data and artificial intelligence technology, this method will be further optimized and improved, providing stronger support for the development of the new materials industry.

[0077] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A new material price trend analysis method based on big data, characterized in that: The specific steps are: S100, Market Data Monitoring: Real-time monitoring and collection of market data, continuous evaluation and optimization of data sources to ensure data quality and diversity, and use data fusion technology to integrate information from different sources to improve data reliability; S200, Material composition impact assessment: Collect composition information from multiple aspects, create a database to store composition data, analyze the impact of composition on material performance and cost, and identify the components that have a significant impact on price; S300, heat coefficient setting: determine the indicators that reflect the market's attention to new materials, collect market heat data through multiple channels, calculate the heat coefficient of each indicator based on the collected data, comprehensively obtain the overall heat coefficient, and establish a dynamic adjustment mechanism for the heat coefficient to adapt to changes in the market environment; S400, price trend forecast: Integrate raw material prices, heat coefficients and market trend data, study and analyze the collected data, establish a price forecast model, verify the forecast results through historical data, and analyze price trends and potential market influencing factors; S500, Decision Support Report Generation: Generate reports based on analysis results, evaluate the uncertainty and potential risks of forecast results, provide procurement and investment recommendations based on analysis results, and present the reports to decision makers in an appropriate format.

2. According to the new material price trend analysis method based on big data as claimed in claim 1, it is characterized in that: The S100 includes: S101, identify key market data indicators and evaluate the reliability, coverage and update frequency of data sources; S102, using web crawler technology to capture data in real time to ensure the timeliness and continuity of data collection; S103, clean the collected data, remove duplications, correct errors, standardize formats, apply data fusion technology, integrate data from different sources, and enhance data consistency; S104, storing the cleaned data in a database and establishing an efficient data index and query mechanism; S105, regularly monitor and evaluate data quality to ensure data accuracy and completeness.

3. According to the new material price trend analysis method based on big data as claimed in claim 1, it is characterized in that: The S200 includes: S201, collect relevant data on ingredients from scientific literature, laboratory reports, and material specifications; S202, creating a relational database based on the ingredient table, performance table, and cost table to store the ingredient data; S203, preprocessing the component data to remove duplicate records, process missing values, and standardize, and merge component data from different sources; S204, using the Pearson correlation coefficient to explore the relationship between composition and performance, and to analyze the impact of changes in composition ratios on material production costs; S205, applying a decision tree to identify key factors affecting the price.

4. According to the new material price trend analysis method based on big data as claimed in claim 3, it is characterized in that: In S204, the relationship between the components and the performance is explored using the Pearson correlation coefficient. Based on the collected component data and performance data, the Pearson correlation coefficient between the components and the performance is calculated. The mathematical expression of the Pearson correlation coefficient is: Where: X i and Y i are the observed values ​​of composition and performance, and is their mean, n is the sample size, r xy The absolute value of close to 1 indicates a strong linear relationship, and close to 0 indicates a weak linear relationship. In order to determine whether the observed correlation is significant, a significance test is performed. The calculation formula of the statistic t of the correlation coefficient is: Among them: r is the correlation coefficient, n is the sample size, and when the value of the statistic is greater than the significance level α, it is considered that there is a significant correlation between the two variables.

5. According to the new material price trend analysis method based on big data as claimed in claim 3, it is characterized in that: In S205, a decision tree is used to identify key components that affect prices, and a data set containing new material prices, material components, and other features that affect prices is collected. The collected data set is cleaned to remove duplicate data, correct erroneous data, and process missing values ​​and outliers. The information gain of each feature is calculated to select features related to prices from the data set. The information gain formula is: Where: D is the data set, A is the attribute, Val(A) is the set of values ​​of attribute A, D v It is to select the data subset after the attribute A value is υ, select the optimal feature for splitting, and divide the data set into different subsets based on the selected optimal feature. Repeat the above feature selection and splitting process for each subset until the stopping condition is met, and a complete decision tree model is constructed. In order to prevent overfitting, it is necessary to prune the constructed decision tree, collect real-time data to evaluate the performance of the decision tree model, analyze the decision tree results, and identify the key components that affect the material price.

6. The new material price trend analysis method based on big data according to claim 1 is characterized in that: The S300 includes: S301, determine the indicators that reflect the market's attention to new materials; S302, collects data from multiple channels including search engine APIs, social media platforms, news websites, industry reports, online markets, and forums; S303, assigning corresponding weights according to the influence of each indicator on market attention, and calculating weighted heat coefficient according to the weights; S304, combining the heat coefficients of the various indicators to obtain an overall heat coefficient; S305, design a dynamic adjustment mechanism to update the heat coefficient based on market feedback and new data; S306, use historical data to verify the effectiveness of the heat coefficient model and make adjustments based on market feedback.

7. The new material price trend analysis method based on big data according to claim 1 is characterized in that: The S400 includes: S401, collects data on raw material prices, heat coefficients, and market trends, integrates data from different sources and types, and builds a comprehensive data set; S402, cleaning the data, processing missing values, outliers and duplicate records, and performing standardization transformation on the cleaned data to eliminate the influence of different dimensions; S403, applying feature selection methods to screen features that have a significant impact on prediction, and dividing the data into a training set and a validation set; S404, selecting a linear regression model to predict future price trends based on data characteristics and prediction targets; S405 uses the training set to train the selected model, adjusts the model parameters, improves the prediction accuracy, and uses the validation set to evaluate the stability and generalization ability of the model using the cross-validation technique; S406, using the trained model to perform price prediction on the latest market data, and generating prediction results of future price trends based on the model output.

8. A new material price trend analysis method based on big data according to claim 7, characterized in that: S404 predicts future price trends by analyzing historical data using a linear regression model, extracting data with significant features into a training set and a validation set, analyzing the distribution characteristics of the data, evaluating the time span and time resolution of the data, determining the predicted time range, and clarifying the application scenarios of the predicted results. The formula of the linear regression model is: y = β0 + β1x1 + β2x2 + ... + β n x n +∈, where: y is the predicted value, x is the raw material price and heat coefficient, β0 is the intercept term, β n is the coefficient, ∈ is the error term, and the model parameters are adjusted by minimizing the loss function using the training set. The mathematical formula for minimizing the sum of squares of the error terms is: Where: m is the number of samples, y is i is the actual value, is the predicted value. The training set is used to adjust the model parameters by minimizing the loss function. The validation set is used to evaluate the prediction performance of the model. The model is adjusted according to the evaluation results, and the trained model is used to predict future prices.

9. The new material price trend analysis method based on big data according to claim 6 is characterized in that: The S500 includes: S501, summarizing the results of the price trend forecasting and market influencing factor analysis steps; S502, using machine learning models to assess the uncertainty of prediction results; S503, identify and assess potential market risks, supply risks, and technical risks; S504, formulate procurement and investment recommendations based on analysis results and risk assessment; S505, integrate analysis results, uncertainty assessment, risk analysis and recommendations into a report.

10. A new material price trend analysis system based on big data, characterized in that: The system includes: Data collection module: responsible for collecting multi-dimensional data such as new material market data, R&D data, and production data, and integrating the data into a unified format and storing it in the database; Data preprocessing module: performs a series of preprocessing operations on the collected data, including cleaning, deduplication, and missing value filling, to ensure the quality and accuracy of the data; Feature extraction module: extracts key features of material type, production process, market demand, and raw material price from preprocessed data as the basis for subsequent prediction and analysis; Prediction model building module: Use deep learning algorithms to create new material price trend prediction models, use historical data and extracted features for training and optimization to improve prediction accuracy; Result display module: The prediction results are displayed to users in a visual form. Users can adjust R&D strategies and market strategies based on the prediction results to improve R&D efficiency and engineering application level.

Images