Mountainous area tobacco inspector demand prediction method

Abstract

A mountainous area tobacco inspector demand prediction method comprises the following steps that X groups of data are collected, including the number of existing inspectors in a current administrative region, the number of certificate holders, the total number of county population, the supervision area, the number of supervision grids, the structure of urban and rural power grid certificate holders, the degree of dispersion of the certificate holders and the regional certificate holding rate; taking the inspector compilation number as a dependent variable and the others as independent variables, performing correlation analysis by using SPSS, and screening the independent variables related to the inspector compilation number; calculating an inspection supervision coefficient J according to the screened variables; clustering the inspection supervision coefficient J by using a K-means clustering analysis method; measuring and calculating quota and staff standards of each category in combination with industry standards and clustering results; the number of inspectors in each region is predicted in combination with an expert scoring method, and scientific prediction and differentiated configuration of inspector demands are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of personnel demand forecasting, and particularly relates to a method for forecasting the demand for tobacco inspectors in mountainous areas. Background Art

[0002] Tobacco inspectors play an important role in the tobacco industry's supervision system. They are mainly responsible for case investigation and intelligence information management in their jurisdiction, maintaining the operation order of the "two types of tobacco" market, and ensuring that the production, sales, and circulation links of tobacco comply with national laws, regulations, and policy requirements. In the tobacco leaf production areas and cigarette sales areas of our country, due to factors such as regional economic development differences, different geographical conditions, wide distribution of tobacco-growing areas, and unbalanced urban and rural distributions, the jurisdiction widths and work difficulties of inspectors vary. The inspection work has strong law enforcement nature, requires certain legal knowledge and law enforcement capabilities, has large regional differences, the characteristics and supervision focuses of the tobacco markets in different regions are different, the inspection work needs to be adjusted to local conditions, has a high work intensity, requires frequent field inspections, and has a complex working environment.

[0003] If the talent recruitment activity is carried out only when there is a shortage of inspectors in the unit, the time required is relatively long, which is likely to affect the unit's production and business activities. Therefore, it is necessary to scientifically allocate the number of inspectors in advance. However, for mountainous area tobacco units, due to significant differences in the administrative areas, urban and rural population distributions, urban and rural network licensee structures, and licensee dispersion degrees of each county or district, the demand for inspectors in each county or district is also different, and it is difficult to draw on the talent recruitment plans of other units. Therefore, how to improve the scientificity and accuracy of the demand for inspectors in the mountainous area tobacco industry is a common technical problem for employers in the current mountainous area tobacco industry. Summary of the Invention

[0004] Based on the above problems, the present invention aims to provide a method for forecasting the demand for tobacco inspectors in mountainous areas, so as to solve the problem of low scientificity and accuracy of the demand for inspectors by employers in the existing mountainous area tobacco industry.

[0005] The technical solution adopted by the present invention is as follows: A method for forecasting the demand for tobacco inspectors in mountainous areas, comprising the following steps:

[0006] Step 1, collect X groups of data of counties (or county-level cities, districts), and each group of data includes the current number of inspector establishment, the number of licensees, the total population, the supervision area, the number of supervision grids, the urban and rural network licensee structure, the licensee dispersion degree, and the regional license rate within the current administrative region;

[0007] Step 2, taking the current number of inspector establishment as the dependent variable, and taking the number of licensees, the total population, the supervision area, the number of supervision grids, the urban and rural network licensee structure, the licensee dispersion degree, and the regional license rate as independent variables, use the SPSS tool to perform a correlation analysis on the dependent variable and the independent variables;

[0008] Step 3: According to the results of the correlation analysis, select the independent variables with a Pearson correlation coefficient greater than 0.5 to obtain n associated variables;

[0009] Step 4: Determine the X groups of inspection and supervision coefficients J based on the n associated variables obtained in Step 3;

[0010] Step 5: Use the K-means clustering analysis method in SPSS to cluster the X groups of inspection and supervision coefficients J to obtain the clustering results;

[0011] Step 6: Based on the industry's inspection staff allocation guidance standards, combined with the clustering analysis results, calculate the corresponding quota standards for each category;

[0012] Step 7: Predict the number of inspectors in each group according to the quota standards.

[0013] Further, the number of licensed households in Step 1 refers to the total number of legal cigarette retail households holding tobacco monopoly retail licenses in the region; the total population refers to the permanent population within the administrative region; the supervision area refers to the total administrative area of a county (or county-level city, district); the number of supervision grids refers to the number of grid-like supervision units into which the supervision area is divided; the urban-rural network licensed household structure refers to the proportional distribution of licensed cigarette retail households in urban and rural areas, urban network licensed household structure = number of urban network licensed households / total number of licensed households in the county-level region * 100%, rural network licensed household structure = number of rural network licensed households / total number of licensed households in the county-level region * 100%; the dispersion degree of licensed households refers to the degree of concentration or dispersion of licensed cigarette retail households in geographical distribution, licensed dispersion degree = total supervision area of the whole county (district) / number of licensed households; the regional licensing rate refers to the ratio of the number of licensed cigarette retail households in a certain region to the permanent population in that region, regional licensing rate = number of regional licensed households / total regional permanent population / 1000 (‰).

[0014] Further, after screening out n associated variables according to the Pearson correlation coefficient in Step 3, use the expert scoring method to score the selected variables to further optimize the variable selection and obtain the final associated variables. Further, the formula for determining the inspection and supervision coefficient J in Step 4 is: ji = associated variable i / average value of associated variable i;

[0015]

[0016] Where J: inspection and supervision coefficient;

[0017] ji: proportion coefficient of associated variable i;

[0018] qi: weight of associated variable i;

[0019] n: number of associated variables;

[0020] The correlation coefficient of the associated variable i is obtained through the correlation analysis in Step 2.

[0021] Further, when using the K-means clustering analysis of SPSS in Step 5, the value of K is set to 3 - 5 categories.

[0022] Further, the guiding standard for the allocation of inspectors in the industry in Step 6 is to allocate one inspector for every 180 households.

[0023] Further, after obtaining the corresponding quota standards for each clustering category in Step 6, the expert scoring method is used to score the corresponding quota standards for each clustering category to determine the final clustering category and the corresponding quota standards.

[0024] Advantages of the present invention: By collecting multiple variables for correlation analysis and through multi-factor comprehensive analysis, the present invention can more accurately capture the differences in the demand for inspectors in different regions, avoiding prediction biases caused by single factors. Prediction through correlation analysis and clustering analysis avoids biases in subjective judgment and has strong scientificity and objectivity. Based on the guiding standard for the allocation of inspectors in the tobacco industry, combined with the results of clustering analysis, the corresponding quota standards for different values of K are calculated; and considering the local actual situation, the expert scoring method is used to determine the final number of inspectors, ensuring that the prediction results are highly relevant to industry policies. The use of the SPSS tool improves the scientificity and reliability of the analysis results of independent variables related to the number of inspectors. The use of the expert scoring method combined with the actual situation enhances the scientificity of decision-making, obtains a more accurate number of inspectors, enables enterprises to plan in advance when recruiting talents, and avoids temporary recruitment from affecting the production and business activities of enterprises. Specific Embodiments

[0025] A method for predicting the demand for tobacco inspectors in mountainous areas includes the following steps:

[0026] Step 1: Collect data of 89 county-level bureaus (branch companies) under the Guizhou Provincial Tobacco Company, including the current number of inspector establishments, the number of licensed households, the total population of the county or district, the supervision area, the number of supervision grids, the urban-rural network licensed household structure, the dispersion degree of licensed households, and the regional licensing rate within the current administrative region;

[0027] Among them:

[0028] The number of licensed households: refers to the total number of legal cigarette retail households holding tobacco monopoly retail licenses in a certain administrative region (including: normal operating households, newly established retail households, and suspended households). It reflects the scale of the cigarette retail market and the legal compliance of retail household business activities and is an important indicator for market supervision.

[0029] Total population of counties and districts: Refers to the permanent population within the administrative region of a certain county (or county-level city, district). It is used to evaluate the potential consumer demand of the cigarette market and the scope of regulatory coverage.

[0030] Regulatory area (county area): Refers to the total administrative area of a certain county (or county-level city, district). It reflects the geographical scope, location, and environment of market supervision, and affects the allocation of regulatory resources and law enforcement efficiency.

[0031] Number of regulatory grids: Refers to the number of grid-based regulatory units into which the county is divided, and each grid is responsible for supervision by a dedicated person. (The county-level area is the first-level grid, towns, townships, sub-districts, economic development zones, etc. are the second-level grids, and communities, village groups, industrial parks, and other areas are the third-level grids) It improves the refinement and efficiency of market supervision, ensuring full coverage and no dead ends.

[0032] Structure of licensed households in urban and rural networks: Refers to the proportion distribution of licensed cigarette retailers in urban and rural areas.

[0033] Structure of licensed households in urban network = Number of licensed households in urban network / Total number of licensed households in the county-level area * 100%

[0034] Structure of licensed households in rural network = Number of licensed households in rural network / Total number of licensed households in the county-level area * 100%

[0035] It reflects the differences between urban and rural markets and helps formulate targeted supervision and service strategies.

[0036] Dispersion degree of licensed households: Refers to the degree of concentration or dispersion of licensed cigarette retailers in geographical distribution.

[0037] Dispersion degree of licensed households = Regulatory area of the whole county (district) / Number of licensed households. A high dispersion degree means that retailers are widely and sparsely distributed, which may increase the difficulty of supervision and distribution; a low dispersion degree means that retailers are concentrated and easy to manage. Regional licensing rate: Refers to the ratio of the number of licensed cigarette retailers in a certain region to the permanent population in that region. Regional licensing rate = Number of licensed households in the region / Total permanent population in the region / 1000 (‰). It reflects the scale of the number of licensed households in a certain region and is an important indicator for measuring whether the "Reasonable Layout Plan for Retail Points of Tobacco Products" is scientific, whether the control of the scale of retail entities is stable, and whether the subsequent supervision of licenses is in place.

[0038] Step 2: Taking the existing number of inspectors in 89 county-level areas in Guizhou Province as the dependent variable, and the number of licensed households, total population of counties, regulatory area (county area), number of regulatory grids, structure of licensed households in urban and rural networks, dispersion degree of licensed households, and regional licensing rate as independent variables, use the SPSS tool to conduct a correlation analysis on the dependent variable and independent variables;

[0039] Table 1 Results of correlation analysis between the number of inspectors and various factors

[0040]

[0041]

[0042] Step 3: Determine the associated variables based on the results of the correlation analysis.

[0043] Through analysis, several factors such as "the number of licensed households, the population of the county / district, the number of supervision grids, the dispersion degree of licensed households, the structure of urban and rural network licensed households, and the regional licensing rate" are all correlated with the number of inspectors to a certain extent and are used as associated variables. Based on the correlation analysis, internal experts in the tobacco monopoly line were organized to conduct a centralized discussion, and it was considered that: A. A large amount of energy and work in tobacco monopoly supervision are focused on key supervised households, and key supervised households are concentrated among urban licensed households. Therefore, to a certain extent, the proportion of urban licensed households can reflect the work difficulty of tobacco monopoly supervision; B. On the one hand, the higher the regional licensing rate, the more concentrated the licensed households, and the higher the supervision efficiency invested per unit time; on the other hand, the higher the dispersion degree of licensed households (supervision area / number of licensed households), the more workload, working hours of market management supervision and tobacco monopoly inspection need to be invested. Therefore, the regional licensing rate and the dispersion degree of licensed households have a greater impact on the setting of the number of inspector positions; C. "Two types of tobacco" counties need to carry out tobacco leaf monopoly inspections during the acquisition period, which has a certain impact on the workload and work difficulty of inspectors.

[0044] Based on the above analysis and research, "the proportion of urban network licensed households", "regional licensing rate", and "dispersion degree of licensed households" are used as associated variables.

[0045] Step 4: Determine the inspection and supervision coefficient J.

[0046]

[0047] ji = associated variable i / average value of associated variable i;

[0048]

[0049] Where J: inspection and supervision coefficient;

[0050] ji: proportion coefficient of associated variable i;

[0051] qi: weight of associated variable i;

[0052] n: number of associated variables;

[0053] According to Step 3, n = 3 is obtained;

[0054] j1: proportion coefficient of urban licensed households;

[0055] j1 = proportion of urban network licensed households in the whole county (district) / average proportion of urban network licensed households in the whole province;

[0056] j2: regional licensed household coefficient;

[0057] j2 = Certificate - holding rate of the whole county (district) / Average certificate - holding rate of the whole province; Regional certificate - holding rate = Number of certificate - holding households in the region / Total regional population * 100%;

[0058] j3: Coefficient of dispersion of certificate - holding households;

[0059] j3 = Dispersion of certificate - holding households in the whole county (district) / Average dispersion of certificate - holding households in the whole province; Certificate - holding dispersion = Regulatory area of the whole county (district) / Number of certificate - holding households;

[0060] Table 2 Calculation results of correlation coefficients and weights of related variables

[0061]

[0062] Finally, we get J = j1 * 40%+j2 * 40%+j3 * 20%;

[0063] Step 5: Using the K - means clustering analysis method of SPSS, cluster the 89 units in the whole province (including Guian New Area Bureau (Company) and 88 county - level bureaus (branch companies)) according to the "Inspection and supervision coefficient", as shown in Table 3, Table 4 and Table 5 specifically.

[0064] ① "Inspection and supervision coefficient" clustering analysis - divided into 3 categories

[0065] Table 3 Clustering analysis table of three - category inspection and supervision coefficients

[0066]

[0067] ② "Inspection and supervision coefficient" clustering analysis - divided into 4 categories

[0068] Table 4 Clustering analysis table of four - category inspection and supervision coefficients

[0069]

[0070]

[0071] Since the maximum number of iteration executions has been reached, the iteration has not converged. The maximum absolute coordinate change of any center is 0.003. The current iteration is 10. The minimum distance between the initial centers is 0.309.

[0072]

[0073] ③ "Inspection and supervision coefficient" clustering analysis - divided into 5 categories

[0074] Table 5 Clustering analysis table of five - category inspection and supervision coefficients

[0075]

[0076] Step 6: Based on the industry's inspector allocation guidance standards and combined with the clustering analysis results, calculate the corresponding quota standards for each category.

[0077] ① Quota standard calculation

[0078] Based on the benchmark of allocating one inspector for every 180 households (industry guidance standard) and combined with the clustering analysis results, calculate the corresponding quota standards for each category.

[0079] Table 6 Calculation table of inspector quota standards

[0080]

[0081]

[0082] Step 7: Determine the final clustering result.

[0083] Based on the above clustering analysis results, organize internal experts in the monopoly line to conduct discussions and analyses. After discussion, it is considered that the clustering result of classifying into 3 categories is more in line with the actual situation of Guizhou Tobacco. However, the classification of individual county-level bureaus (branch companies) should be slightly adjusted, and the quota standards should be rounded according to the arithmetic progression increasing rule.

[0084] Table 7 Final inspector quota standards

[0085]

[0086]

[0087]

[0088] Predict the number of inspectors according to the final clustering result.

[0089] Table 8 Predicted number of inspectors in county-level bureaus (branch companies) of Qiannan Prefecture, Guizhou Province

[0090]

[0091] The above content is a further detailed description of the present invention in combination with specific preferred implementation manners. It cannot be determined that the specific implementation of the present invention is only limited to these descriptions. For those of ordinary skill in the technical field to which the present invention belongs, without departing from the concept of the present invention, several simple deductions or substitutions can still be made, which should all be regarded as belonging to the protection scope of the present invention.

Claims

1. A method for predicting demand for tobacco inspectors in mountainous areas, characterized in that: The following steps are involved: Step 1: Collect data from X groups of counties (or county-level cities, districts), each of which includes the number of existing inspectors, the number of licensed households, the total population, the supervision area, the number of supervision grids, the structure of licensed households in urban and rural areas, the dispersion of licensed households, and the regional license rate in the current administrative area; Step 2: With the number of existing auditors as the dependent variable, the number of licensed households, total population, supervision area, number of supervision grids, structure of licensed households in urban and rural areas, dispersion of licensed households and regional licensed rate as independent variables, SPSS tools were used to conduct correlation analysis on the dependent and independent variables; Step 3: According to the correlation analysis results, select the independent variables with a Pearson correlation coefficient greater than 0.5 to obtain n associated variables; Step 4: Determine the audit supervision coefficient J of group X based on the n associated variables obtained in step 3; Step 5: Use the K-means cluster analysis method of SPSS to cluster the audit supervision coefficient J of group X and obtain the clustering results; Step 6: Based on the guidance standards for auditor allocation in the industry and combined with the cluster analysis results, calculate the corresponding quota staffing standards for each category; Step seven: Predict the number of auditors in each group based on the quota staffing standards.

2. A method for predicting demand for tobacco inspectors in mountainous areas according to claim 1, characterized in that: In the step 1, the number of licensed households refers to the total number of legal cigarette retailers holding tobacco monopoly retail licenses in the region; the total population refers to the number of permanent residents within the administrative area; the supervision area refers to the total administrative area of ​​the county (or county-level city, district); the number of supervision grids refers to the number of grid-based supervision units into which the supervision area is divided; the structure of licensed households in urban and rural areas refers to the proportional distribution of licensed cigarette retailers in urban and rural areas, the structure of licensed households in urban areas = the number of licensed households in urban areas / the total number of licensed households in county-level areas*100%, the structure of licensed households in rural areas = the number of licensed households in rural areas / the total number of licensed households in county-level areas*100%; the dispersion of licensed households refers to the degree of concentration or dispersion of licensed cigarette retailers in geographical distribution, the dispersion of license = the supervision area of ​​the entire county (district) / the number of licensed households; the regional license rate refers to the ratio of the number of licensed cigarette retailers in a certain area to the number of permanent residents in that area, the regional license rate = the number of licensed households in the area / the total number of permanent residents in the area / 1000 (‰).

3. A method for predicting demand for tobacco inspectors in mountainous areas according to claim 1, characterized in that: In the step three, after n associated variables are screened out according to the Pearson correlation coefficient, the screened variables are scored using the expert scoring method to further optimize the variable selection and obtain the final associated variables.

4. A method for predicting demand for tobacco inspectors in mountainous areas according to claim 1, characterized in that: The formula for determining the audit supervision coefficient J in step 4 is: ji = associated variable i / average value of associated variable i; Where J: audit supervision coefficient; ji: coefficient of association variable i; qi: weight of associated variable i; n: the number of associated variables; The correlation coefficient of associated variable i is obtained by the correlation analysis in step 2.

5. A method for predicting demand for tobacco inspectors in mountainous areas according to claim 1, characterized in that: When using the K-means cluster analysis of SPSS in step 5, the K value is set to 3-5 categories.

6. A method for predicting demand for tobacco inspectors in mountainous areas according to claim 1, characterized in that: The guidance standard for allocating auditors in the industry described in step six is ​​one auditor for every 180 households.

7. A method for predicting demand for tobacco inspectors in mountainous areas according to claim 1, characterized in that: After obtaining the quota and staffing standards corresponding to each cluster category in step 6, the expert scoring method is used to score the quota and staffing standards corresponding to each cluster category to determine the final cluster category and the corresponding quota and staffing standards.

Images