Material purchasing recommendation method and system based on statistical learning

Abstract

The invention discloses a statistical learning-based material purchasing recommendation method and system, which is applied to the technical field of material purchasing management, and comprises the following steps: carrying out attribute decomposition on the price of a material to be purchased, constructing an expected budget model, and calculating a budget expected value; and through the expected budget value, matching the lowest price from the published information of the suppliers, and carrying out material purchasing. According to the invention, the cost performance of the materials is comprehensively known, so that the cost increase of the purchase cost of the materials can be avoided, the efficiency of purchase management is effectively improved, a user can select the most adaptive supplier and push a certain number of suppliers with high matching degree to a purchase publisher, and an enterprise can be enabled not to depend on the actual combat experience of the purchase personnel, so that the purchase management efficiency is improved. Projects are accurately and efficiently purchased and completed, and profits are obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of material procurement management, and more specifically to a material procurement recommendation method and system based on statistical learning. Background Art

[0002] With the development of the times, technology has gradually spread to every corner of people's lives, among which material procurement management technology is a specific manifestation. It better guarantees the purchase of users' materials, helps users experience more favorable and labor-saving procurement, and allows users to buy the highest value configuration at the lowest cost. At the same time, it also saves manpower and time to a certain extent, making their purchases more efficient and secure, effectively managing purchase costs, and avoiding budget overspending and waste.

[0003] The current technology for purchasing and managing power materials mainly relies on manual methods. This manual purchasing method is time-consuming and labor-intensive. In addition, manual selection and comparison of materials cannot fully understand the cost-effectiveness of power materials, which will cause overspending on the purchase cost of power materials. In addition, the efficiency of purchasing management cannot be effectively improved, and users cannot choose the most suitable supplier.

[0004] Therefore, how to provide a material procurement recommendation method is an urgent problem that technical personnel in this field need to solve. Summary of the invention

[0005] In view of this, the present invention provides a material procurement recommendation method and system based on statistical learning to solve the problems in the background technology.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] The present invention discloses a material procurement recommendation method based on statistical learning, comprising:

[0008] Decompose the prices of materials to be purchased into raw material cost data, fixed cost data, and variable cost data;

[0009] Determine the amount of main raw materials according to product process standards, and determine the raw material cost data of materials to be purchased based on raw material prices;

[0010] Collect product cost information from multiple suppliers, use statistical learning methods to eliminate abnormal data, estimate data distribution, and determine the fixed cost data of products;

[0011] Collect variable cost data from multiple suppliers, calculate the output cost rate of each variable cost for each supplier, and then combine statistical learning to obtain the estimated output cost rate of each variable cost;

[0012] Construct an expected budget model and calculate the expected budget value based on raw material cost data, fixed cost data, and variable cost data;

[0013] The expected budget is used to match the lowest price from the information published by multiple suppliers to purchase materials.

[0014] Preferably, in the above-mentioned material procurement recommendation method based on statistical learning, the specific steps of determining the fixed cost data of the product are as follows:

[0015] Eliminate outliers. Determine whether the product cost information of multiple suppliers is normally distributed. If it is normally distributed, use the Laida criterion and the Romanovsky criterion to eliminate outliers. If it is not normally distributed, use the box plot method and the isolation forest algorithm to eliminate outliers.

[0016] Estimate data distribution, calculate confidence intervals, and use moving confidence interval method, median confidence interval method, and Bootstrap interval method to determine the range of main raw material usage;

[0017] Select max{lower bound of moving confidence interval method, lower bound of median confidence interval method, lower bound of Bootstrap method} as the lower bound of the minimum confidence interval, and select min{upper bound of moving confidence interval method, upper bound of median confidence interval method, upper bound of Bootstrap method} as the upper bound of the minimum confidence interval;

[0018] Determine the fixed cost data for the product.

[0019] Preferably, in the above-mentioned material procurement recommendation method based on statistical learning, the specific steps for determining the output value cost rate of each variable cost are as follows: obtain variable cost data from the supplier's financial statement data; the variable cost data include: labor costs, equipment depreciation, transportation costs, administrative expenses, sales expenses, financial expenses, taxes, calculate the proportion of each variable cost data to the main business income, and determine the output value cost rate of each variable cost.

[0020] Preferably, in the above-mentioned material procurement recommendation method based on statistical learning, the expected budget model formula is as follows:

[0021]

[0022] Where y: budget expectation, C1: raw material cost data, C2: fixed cost data, c d : Equipment depreciation rate, c t : Transportation cost ratio, c h : Labor cost ratio, c m : Management expense ratio, c s : Transportation cost ratio, cf : Labor cost ratio, c it : Income tax rate, c a : Tax and surcharge ratio, r is the profit margin, which is determined by the supplier.

[0023] Preferably, in the above-mentioned material procurement recommendation method based on statistical learning, in the step of determining the fixed cost data of the product, the moving confidence interval method adopts a method of traversing the search interval, traversing all confidence intervals, and finding all intervals that can satisfy the sample size ratio of μ (50% ≤ μ ≤ 100%), and taking the shortest interval as the confidence interval of the data.

[0024] Preferably, in the above-mentioned material procurement recommendation method based on statistical learning, in the step of determining the fixed cost data of the product, the median confidence interval method includes:

[0025] (1) If the sample size is less than 20, the upper and lower limits are calculated as follows:

[0026]

[0027] In formulas (1) and (2), n represents the sample size; Z α Determined by the confidence level; two percentiles are calculated according to formulas (1) and (2), and multiplied by the sample size n to obtain the ranking of the upper and lower limits of the confidence interval of the population median; then the original data are sorted from small to large to obtain the confidence interval of the median; or the confidence interval of the median is calculated using the percentile calculation formula;

[0028] (2) If the sample size is greater than 20, the formula for calculating the rank of the upper and lower limits of the 1-α confidence interval of the population median is:

[0029]

[0030] Among them, Z α is the two-sided limit of the standard normal distribution; S1 and S2 represent the rank of the upper and lower limits of the confidence interval, respectively. Here S1 is S * Round off; sort the original data from small to large; and obtain the confidence interval of the overall median.

[0031] Preferably, in the above-mentioned material procurement recommendation method based on statistical learning, in the step of determining the fixed cost data of the product, the specific steps of the Bootstrap interval method include:

[0032] (4) In a data set containing m samples, randomly select one sample each time, use it as a training sample, and then put this sample back into the data set;

[0033] (5) Sampling with replacement is done m times to generate a data set of the same size as the original data set. This new data set is the training set.

[0034] (6) Use the samples that were not selected to construct a validation set and calculate the validation error of the model.

[0035] The present invention discloses a material procurement recommendation system based on statistical learning, which applies the above-mentioned recommendation method and includes a procurement end and a supplier end; the procurement end collects and stores original data collected from multiple supplier ends, performs preprocessing, and builds an expected budget model to obtain the expected budget value; compares the data with the price data released by the multiple supply ends, and matches the selected supplier.

[0036] Preferably, in the above-mentioned material procurement recommendation system based on statistical learning, the procurement end includes:

[0037] A data acquisition module is used to collect and store raw data collected from various channels, and pre-process the raw data to obtain raw material cost data, fixed cost data, and variable cost data;

[0038] The expected budget model building module is used for the statistical calculation of cost data, including iterative calculation of raw material costs, statistical calculation of fixed costs and statistical calculation of variable costs, to obtain the expected budget value;

[0039] The recommendation matching module matches the supplier with the lowest price from the information published by multiple suppliers according to the expected budget value.

[0040] Preferably, in the above-mentioned material procurement recommendation system based on statistical learning, the supplier side publishes price and product information; and establishes data connection with the procurement side after matching is completed.

[0041] It can be seen from the above technical solution that compared with the prior art, the present invention discloses a material procurement recommendation method and system based on statistical learning, constructs an expected budget model for data analysis and processing, and is mainly used for the enterprise's material procurement price control, analyzes, identifies and outputs the prices of different suppliers of the enterprise, reduces the economic losses caused to the company by high-priced procurement, and comprehensively understands the cost-effectiveness of materials, thereby avoiding overspending on material purchase costs, and the efficiency of procurement management is effectively improved. Users can also choose the most suitable supplier and push a certain number of highly matched suppliers to the procurement publisher, so that the enterprise can accurately and efficiently purchase and complete the project without relying on the practical experience of procurement personnel to obtain profits. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] 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 description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0043] Figure 1 A flow chart provided for the present invention;

[0044] Figure 2 It is a data processing flow chart of the present invention. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. 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.

[0046] Embodiment 1 of the present invention discloses and provides a material procurement recommendation method based on statistical learning, and the specific steps are as follows:

[0047] Decompose the prices of materials to be purchased into raw material cost data, fixed cost data, and variable cost data;

[0048] Determine the amount of main raw materials according to product process standards, and determine the raw material cost data of materials to be purchased based on raw material prices;

[0049] Collect product cost information from multiple suppliers, use statistical learning methods to eliminate abnormal data, estimate data distribution, and determine the fixed cost data of products;

[0050] Collect variable cost data from multiple suppliers, calculate the output cost rate of each variable cost for each supplier, and then combine statistical learning to obtain the estimated output cost rate of each variable cost;

[0051] Construct an expected budget model and calculate the expected budget value based on raw material cost data, fixed cost data, and variable cost data;

[0052] The expected budget is used to match the lowest price from the information published by multiple suppliers to purchase materials.

[0053] Furthermore, the steps of the method of the present invention are described in conjunction with specific embodiments.

[0054] Embodiment 1:

[0055] A domestic enterprise needs to make a price forecast for cement poles to assist in its raw material bidding decision. The 190-10I cement pole model was selected, and the original data was collected and desensitized, including standard drawing data, supplier's financial statements in recent years, supplier survey receipts, and national enterprise procurement transaction sourcing inquiry system data. The specific operation steps are as follows:

[0056] Step ①: Calculation of reinforced concrete cost (raw material cost);

[0057] The consumption data of steel bars and concrete for three types of cement poles were collected through standard drawings for cement pole production. The standard consumption of steel bars and concrete for the whole pole group and flange group of the same type of cement pole is the same.

[0058] The prices of corresponding types of reinforced concrete can be queried through the National Enterprise Procurement Transaction Sourcing and Price Inquiry System, as shown in Table 1.

[0059] Table 1 Calculation of reinforced concrete prices

[0060]

[0061] By formula

[0062] Steel bar / concrete = standard quantity × unit price;

[0063] The price of raw materials for reinforced concrete can be obtained.

[0064] Step 2: Determine fixed costs;

[0065] The data on the quantity and unit price of raw materials (including admixtures, flanges, clamps, spiral reinforcement (erection rings), auxiliary materials, equipment maintenance, excess propeller treatment, flange bolts, water charges, electricity, and steam) are extracted from the standard quotation sheets of cement pole manufacturers.

[0066] Calculate the product of unit price and quantity of fixed cost of raw materials, as shown in Table 2.

[0067] Table 2 Raw material costs

[0068]

[0069] Secondly, the Romanovsky criterion, Laida criterion and box plot method were used to eliminate abnormal data, as shown in Table 3;

[0070] Table 3 Fixed cost after outliers are removed

[0071]

[0072] After removing abnormal data, the confidence interval is calculated using the moving confidence interval method, the median confidence interval method, and the Bootstrap method, including the maximum value, minimum value, and median of the confidence interval, as shown in Table 4.

[0073] Table 4 Results of three confidence interval methods

[0074]

[0075] After comparison, the minimum confidence interval was obtained, and the estimated fixed cost of this model of cement pole is 11.90.

[0076] Step ③: Determine the accounting of various operating expenses;

[0077] The output value cost rate method is used to calculate the proportion of various costs to main operating income. The calculation formula is:

[0078]

[0079] Similarly, first determine whether the output cost rate data of each cost obeys the normal distribution. For data that obeys the normal distribution, use the T test or the Laida criterion to eliminate abnormal data; for data that does not obey the normal distribution, use the box plot method to process the data. Then calculate the confidence interval, use the moving confidence interval method, the median confidence interval method and the Bootstrap method to calculate the confidence interval, and finally take the median of the minimum interval, as shown in Table 5.

[0080] Table 5 Variable cost coefficients

[0081]

[0082] Step ④: Determine the purchase price of cement poles;

[0083] The following core formula:

[0084]

[0085] Determine the price of this type of cement pole. Among them, y: product purchase price, C1: raw material cost, C2: fixed cost, c d : Equipment depreciation rate, c t : Transportation cost ratio, c h : Labor cost ratio, c m : Management expense ratio, c s : Transportation cost ratio, c f : Labor cost ratio, c it : Income tax rate, c a : Tax and surcharge ratio, r is the profit margin.

[0086] The error between the predicted price and the actual selling price of cement poles is calculated, as shown in Table 6.

[0087] Table 6 Calculation results

[0088]

[0089] Embodiment 2:

[0090] Select another cement pole model 190-12M and use the same method to predict its price. The specific steps are as follows:

[0091] Step ①: Calculation of reinforced concrete cost (raw material cost);

[0092] The consumption data of steel bars and concrete for three types of cement poles were collected through standard drawings for cement pole production. The standard consumption of steel bars and concrete for the whole pole group and flange group of the same type of cement pole is the same.

[0093] The prices of corresponding types of reinforced concrete can be checked through the National Enterprise Procurement Transaction Sourcing and Inquiry System, as shown in Table 7.

[0094] Table 7 Calculation of reinforced concrete prices

[0095]

[0096]

[0097] By the formula:

[0098] Steel bar / concrete = standard quantity × unit price;

[0099] The price of raw materials for reinforced concrete can be obtained.

[0100] Step 2: Determine fixed costs;

[0101] The data on the quantity and unit price of raw materials (including admixtures, flanges, clamps, spiral reinforcement (erection rings), auxiliary materials, equipment maintenance, excess propeller treatment, flange bolts, water charges, electricity, and steam) are extracted from the standard quotation sheets of cement pole manufacturers.

[0102] Calculate the product of unit price and quantity of raw material cost, as shown in Table 8.

[0103]

[0104] Secondly, the Romanovsky criterion, Laida criterion and box plot method were used to eliminate abnormal data, as shown in Table 9;

[0105] Table 9 After removing outliers in fixed costs

[0106]

[0107]

[0108] After removing abnormal data, the confidence interval was calculated using the moving confidence interval method, the median confidence interval method, and the Bootstrap method, including the maximum value, minimum value, and median of the confidence interval, as shown in Table 10.

[0109] Table 10 Results of three confidence interval methods

[0110]

[0111] After comparison, the minimum confidence interval was obtained, and the fixed cost of this model of cement pole was estimated to be 14.99.

[0112] Step ③: Determine the accounting of various operating expenses;

[0113] The output value cost rate method is used to calculate the proportion of various costs to main operating income. The calculation formula is:

[0114]

[0115] Similarly, first determine whether the output value cost rate data of each cost obeys the normal distribution. For data that obeys the normal distribution, use the T test or the Laida criterion to eliminate abnormal data; for data that does not obey the normal distribution, use the box plot method to process the data. Then calculate the confidence interval, use the moving confidence interval method, the median confidence interval method and the Bootstrap method to calculate the confidence interval, and finally take the median of the minimum interval, as shown in Table 11.

[0116] Table 11 Variable cost coefficients

[0117]

[0118]

[0119] Step ④: Determine the price of cement poles;

[0120] The following core formula:

[0121]

[0122] (where y: product purchase price, C1: raw material cost, C2: fixed cost, c d : Equipment depreciation rate, c t : Transportation cost ratio, c h : Labor cost ratio, c m : Management expense ratio, c s : Transportation cost ratio, c f : Labor cost ratio, c it : Income tax rate, c a : Tax and surcharge ratio, r is the profit margin, which is determined by the supplier)

[0123] Determine the purchase price of this model of cement pole.

[0124] The error between the predicted price and the actual selling price of cement poles is calculated, as shown in Table 12.

[0125] Table 12 Calculation results

[0126]

[0127] Embodiment three:

[0128] Select the insulator model R8ET170L and use the same method to predict its price. The specific steps are as follows:

[0129] Step ①: Calculation of the cost of clay, feldspar powder and quartz powder (raw material cost);

[0130] The ratio of the three raw materials (5:3:2) was determined through research, and the amount of the three main materials can be determined by the weight of the insulator, as shown in Table 13.

[0131]

[0132] The prices of corresponding types of reinforced concrete can be checked through the National Enterprise Procurement Transaction Sourcing and Inquiry System, as shown in Table 14.

[0133] Table 14 Calculation of reinforced concrete prices

[0134]

[0135] By the formula:

[0136] Clay / feldspar powder / quartz powder = standard dosage × unit price;

[0137] The raw material price of the insulator can be obtained.

[0138] Step 2: Determine fixed costs;

[0139] By analyzing the manufacturer's survey letter, the amount and price of auxiliary materials such as glaze and steel feet can be extracted; in addition, the price range of various auxiliary materials can be obtained on the procurement website. After statistical analysis such as normality test, abnormal data elimination, and confidence interval judgment on these data, the reasonable average fixed cost of the insulator model is obtained, as shown in Table 15.

[0140] Table 15 Fixed cost of insulator raw materials

[0141]

[0142] Step 3: Calculate variable costs;

[0143] The output value cost rate method is used to calculate the proportion of various costs to main operating income. The calculation formula is:

[0144]

[0145] Similarly, first determine whether the output cost rate data of each cost obeys the normal distribution. For data that obeys the normal distribution, use the T test or the Laida criterion to eliminate abnormal data; for data that does not obey the normal distribution, use the box plot method to process the data. Then calculate the confidence interval, use the moving confidence interval method, the median confidence interval method and the Bootstrap method to calculate the confidence interval, and finally take the median of the minimum interval, as shown in Table 16.

[0146] Table 16 Variable cost coefficients

[0147]

[0148] Step ④: Determine the purchase price of insulators

[0149] The following core formula:

[0150]

[0151] (where y: product purchase price, C1: raw material cost, C2: fixed cost, c d : Equipment depreciation rate, c t : Transportation cost ratio, c h : Labor cost ratio, c m : Management expense ratio, c s : Transportation cost ratio, c f : Labor cost ratio, c it : Income tax rate, c a : Tax and surcharge ratio, r is the profit margin, which is determined by the supplier)

[0152] Determine the price of this model of insulator.

[0153] The error between the predicted price and the actual selling price of the insulator is calculated, as shown in Table 17.

[0154] Table 17 Calculation results

[0155]

[0156] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0157] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A material procurement recommendation method based on statistical learning, characterized in that: include: Decompose the prices of materials to be purchased into raw material cost data, fixed cost data, and variable cost data; Determine the amount of main raw materials according to product process standards, and determine the raw material cost data of materials to be purchased based on raw material prices; Collect product cost information from multiple suppliers, use statistical learning methods to eliminate abnormal data, estimate data distribution, and determine the fixed cost data of products; Collect variable cost data from multiple suppliers, calculate the output cost rate of each variable cost for each supplier, and then combine statistical learning to obtain the estimated output cost rate of each variable cost; Construct an expected budget model and calculate the expected budget value based on raw material cost data, fixed cost data, and variable cost data; The expected budget is used to match the lowest price from the information published by multiple suppliers to purchase materials.

2. A material procurement recommendation method based on statistical learning according to claim 1, characterized in that: The specific steps to determine the fixed cost data of a product are as follows: Eliminate outliers. Determine whether the product cost information of multiple suppliers is normally distributed. If it is normally distributed, use the Laida criterion and the Romanovsky criterion to eliminate outliers. If it is not normally distributed, use the box plot method and the isolation forest algorithm to eliminate outliers. Estimate data distribution, calculate confidence intervals, and use moving confidence interval method, median confidence interval method, and Bootstrap interval method to determine the range of main raw material usage; Select max{lower bound of moving confidence interval method, lower bound of median confidence interval method, lower bound of Bootstrap method} as the lower bound of the minimum confidence interval, and select min{upper bound of moving confidence interval method, upper bound of median confidence interval method, upper bound of Bootstrap method} as the upper bound of the minimum confidence interval; Determine the fixed cost data for the product.

3. The material procurement recommendation method based on statistical learning according to claim 1 is characterized in that: The specific steps to determine the output cost rate of each variable cost are as follows: obtain variable cost data from the supplier's financial statement data; The variable cost data include: labor costs, equipment depreciation, transportation costs, management costs, sales costs, financial costs, taxes, calculate the proportion of each variable cost data to the main business income, and determine the output value cost rate of each variable cost.

4. The material procurement recommendation method based on statistical learning according to claim 1 is characterized in that: The expected budget model formula is as follows: Where y: budget expectation, C1: raw material cost data, C2: fixed cost data, c d : Equipment depreciation rate, c t : Transportation cost ratio, c h : Labor cost ratio, c m : Management expense ratio, c s : Transportation cost ratio, c f : Labor cost ratio, c it : Income tax rate, c a : Tax and surcharge ratio, r is the profit margin, which is determined by the supplier.

5. The material procurement recommendation method based on statistical learning according to claim 2 is characterized in that: In the step of determining the fixed cost data of the product, the moving confidence interval method adopts the method of traversing the search interval, traversing all confidence intervals, finding all intervals that can satisfy the sample size ratio of μ (50% ≤ μ ≤ 100%), and taking the shortest interval as the confidence interval of the data.

6. The material procurement recommendation method based on statistical learning according to claim 2 is characterized in that: In the step of determining the fixed cost data of a product, the median confidence interval method includes: (1) If the sample size is less than 20, the upper and lower limits are calculated as follows: In formulas (1) and (2), n represents the sample size; Z α Determined by the confidence level; two percentiles are calculated according to formulas (1) and (2), and multiplied by the sample size n to obtain the ranking of the upper and lower limits of the confidence interval of the population median; then the original data are sorted from small to large to obtain the confidence interval of the median; or the confidence interval of the median is calculated using the percentile calculation formula; (2) If the sample size is greater than 20, the formula for calculating the rank of the upper and lower limits of the 1-α confidence interval of the population median is: Among them, Z α is the two-sided limit of the standard normal distribution; S1 and S2 represent the rank of the upper and lower limits of the confidence interval, respectively. Here S1 is S * Round off; sort the original data from small to large; and obtain the confidence interval of the overall median.

7. The material procurement recommendation method based on statistical learning according to claim 2 is characterized in that: In the step of determining the fixed cost data of the product, the specific steps of the Bootstrap interval method include: (1) In a data set containing m samples, randomly select one sample each time, use it as a training sample, and then put this sample back into the data set; (2) Sampling with replacement is performed m times to generate a dataset of the same size as the original dataset. This new dataset is the training set. (3) Use the samples that were not selected to construct a validation set and calculate the validation error of the model.

8. A material procurement recommendation system based on statistical learning, applying the material procurement recommendation method based on statistical learning described in any one of claims 1 to 7, characterized in that: It includes a purchasing side and a supplier side; the purchasing side collects and stores raw data collected from multiple supplier sides, performs preprocessing, and builds an expected budget model to obtain the expected budget value; compares the price data released by multiple supply sides, and matches with the selected supplier.

9. The material procurement recommendation system based on statistical learning according to claim 8, characterized in that: The purchasing end includes: A data acquisition module is used to collect and store raw data collected from various channels, and pre-process the raw data to obtain raw material cost data, fixed cost data, and variable cost data; The expected budget model building module is used for the statistical calculation of cost data, including iterative calculation of raw material costs, statistical calculation of fixed costs and statistical calculation of variable costs, to obtain the expected budget value; The recommendation matching module matches the supplier with the lowest price from the information published by multiple suppliers according to the expected budget value.

10. A material procurement recommendation system based on statistical learning according to claim 8, characterized in that: The supplier side publishes price and product information; and after matching is completed, a data connection is established with the purchase side.

Images