CN116016303A - Method for identifying service quality problem of core network based on artificial intelligence - Google Patents
Method for identifying service quality problem of core network based on artificial intelligence Download PDFInfo
- Publication number
- CN116016303A CN116016303A CN202211545658.1A CN202211545658A CN116016303A CN 116016303 A CN116016303 A CN 116016303A CN 202211545658 A CN202211545658 A CN 202211545658A CN 116016303 A CN116016303 A CN 116016303A
- Authority
- CN
- China
- Prior art keywords
- data
- information entropy
- attribute
- value
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 11
- 238000003066 decision tree Methods 0.000 claims description 17
- 238000012545 processing Methods 0.000 claims description 10
- 230000015556 catabolic process Effects 0.000 claims description 9
- 238000006731 degradation reaction Methods 0.000 claims description 9
- 238000013459 approach Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 5
- 238000012217 deletion Methods 0.000 claims description 3
- 230000037430 deletion Effects 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000011002 quantification Methods 0.000 claims description 2
- 239000000758 substrate Substances 0.000 claims 5
- 239000000523 sample Substances 0.000 claims 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
Abstract
The invention provides a core network service quality problem identification method based on artificial intelligence, which belongs to the technical field of network quality identification. The method predicts and identifies the related business problems of the core network and obtains excellent results.
Description
Technical Field
The invention relates to the technical field of network quality identification, in particular to a core network service quality problem identification method based on artificial intelligence.
Background
With the popularization of internet technology, the demand for the internet is increasing. The quality of internet access is receiving more widespread attention as an important indicator for measuring user experience. The traditional hard coding mode can not efficiently and accurately identify the degradation state of the core network, so that huge hidden danger of operation and maintenance of the core network is generated, and related experience of a user is influenced.
Disclosure of Invention
In order to solve the technical problems, the invention provides an artificial intelligence-based core network service quality problem identification method. Based on the background, the thought and the method of artificial intelligence are adopted to predict and identify the related service problems of the core network, and excellent results are obtained.
The technical scheme of the invention is as follows:
a core network service quality problem identification method based on artificial intelligence constructs a decision tree according to relevant characteristic data in service and models, judges the degradation of service data through an optimization framework LGBM of the decision tree, and predicts the degradation of subsequent service data.
Further, the method comprises the steps of,
firstly, data preprocessing:
modeling is carried out by taking the operation time as a demarcation point and adopting data with a time period of one month, wherein the granularity of the data is 5 minutes. And judging whether the service is deteriorated or not through the prediction of the data.
Null value processing:
1.1 deleting characteristics containing missing values: if the deletion rate of the variable is more than 80%, directly deleting the variable;
1.2 mean value interpolation; the attribute of the data is divided into a fixed-distance type and a non-fixed-distance type, if the missing value is fixed-distance type, the missing value is interpolated by the average value of the existence value of the attribute; if the missing values are non-fixed, the missing values are complemented with the mode of the attribute (i.e., the value with the highest frequency of occurrence) according to the mode principle in statistics.
Outlier processing:
absolute value difference median method: the method of detecting outliers by summing the distances between all factors and the average value is first required. Logic for processing:
step one, finding out the median Xmedia of all factors;
step two, obtaining an absolute deviation value Xi-Xmedia of each factor and a median;
thirdly, obtaining a median MAD of the absolute deviation value; finally, the parameter n is determined, so that a reasonable range is determined as [ Xmedian-nMAD, xmedian+nmad ], and an adjustment is made for the factor values out of the reasonable range.
Further, the method comprises the steps of,
and the optimization framework LGBM based on the decision tree carries out identification prediction on the service problems. Searching an attribute field with the maximum information quantity in a database by utilizing the information gain in the information theory, establishing a node of a decision tree, establishing branches of the tree according to different values of the attribute field, and repeatedly establishing lower nodes and branches of the tree in each branch subset.
The concrete process of constructing the decision tree is as follows:
first find the initial split, the whole training set is used as the set for generating decision tree, each record of the training set must be classified to decide which attribute domain is the best classification index at present,
firstly, exhausting all attribute domains, quantifying the quality of splitting of each attribute domain, and calculating the best splitting, wherein the quantification standard is to calculate the diversity index of each splitting; next, the previous step is repeated until the records within each leaf node belong to the same class and grow to a complete tree.
First select bifurcation feature approach: based on the information gain of the information entropy, equation 1 represents the degree of confusion of the data, and may define the information entropy:
when there is only one data category in equation 1, i.e., p (x i )＝1，log 2 p(x i ) Because h=0, the purer the data, the more categories H is greater;
then the information gain needs to be defined as shown in equation 2; the information entropy of the child node is shown in a formula 3;
Gain＝H(D)-H′(D)
equation 2
The information entropy of the father node and the information entropy of the child node are different, and the information entropy of the father node is called as 'category information entropy'; the information entropy of the child node is called as "condition information entropy" or "attribute information entropy"; information gain = category information entropy-attribute information entropy; when the information gain result is larger than and approaches to 1, the label is 1, and the index degradation is represented in the service, and vice versa.
The invention has the beneficial effects that
The invention can identify the service quality problem of the core network related service VoNR, voLTE, EPSFB and can predict the complaint probability of the 5G user. The method is applied to the actual production environment, and currently, the whole network calculation is carried out on the sampling data of the single address user in a certain province, and the corresponding prediction result is output. And the accuracy rates of 100%, 85% and 97.5% are obtained in the corresponding services, so that a more accurate and efficient method is provided for identifying the service problems of the core network.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments, and all other embodiments obtained by persons of ordinary skill in the art without making any inventive effort based on the embodiments of the present invention are all within the scope of protection of the present invention.
The invention discloses a core network service quality problem identification method based on artificial intelligence, which is used for judging whether the service quality of a core network is deteriorated after network operation is completed. Wherein the core network comprises VoNR, voLTE, EPSFB and 5G customer network entry complaint predictions. And (3) completing mathematical modeling through service history data, and predicting related service problems of the core network in actual application.
The invention analyzes the data of the service problem of the core network based on the artificial intelligence method, is applied to the analysis and prediction of the actual service problem, comprehensively considers and improves the prediction accuracy of the service problem of the core network and improves the working efficiency of related services, so that the method for identifying the service quality problem of the core network based on the artificial intelligence is developed.
The specific implementation process of the invention is as follows:
a) Data preprocessing:
modeling is carried out by taking the operation time as a demarcation point and adopting data with a time period of one month, wherein the granularity of the data is 5 minutes. And judging whether the service is deteriorated or not through the prediction of the data.
1. Null value processing:
1.1 deleting characteristics containing missing values: if the deletion rate of the variable is higher (more than 80%), the coverage rate is lower, and the importance is lower, the variable can be directly deleted.
1.2 mean interpolation. The attributes of the data are classified into a fixed-distance type and a non-fixed-distance type. If the missing value is a fixed distance value, interpolating the missing value with the average value of the attribute presence values; if the missing value is of the non-fixed type, the missing value is complemented by the mode of the attribute (i.e. the value with the highest frequency of occurrence) according to the mode principle in statistics
2 outlier processing:
2.1 median absolute value difference method: the method of detecting outliers by summing the distances between all factors and the average value is first required.
Logic for processing:
step one, finding out the median Xmedia of all factors;
step two, obtaining an absolute deviation value Xi-Xmedia of each factor and a median;
thirdly, obtaining a median MAD of the absolute deviation value; finally, the parameter n is determined, so that a reasonable range is determined as [ Xmedian-nMAD, xmedian+nmad ], and an adjustment is made for the factor values out of the reasonable range.
b) Algorithm implementation
The algorithm is mainly based on an optimization framework LGBM of a decision tree to identify and predict service problems. The decision tree method is to find the attribute field with the maximum information in the database by using the information gain in the information theory, establish a node of the decision tree, establish branches of the tree according to different values of the attribute field, and repeatedly establish a process of the lower node and the branches of the tree in each branch subset, wherein the concrete process of constructing the decision tree is as follows:
firstly, searching an initial split, taking the whole training set as a set for generating a decision tree, wherein each record of the training set is classified to determine which attribute domain is the best classification index at present, generally, exhausting all attribute domains, quantifying the split of each attribute domain, calculating the best split, calculating the quantized standard to calculate the diversity index of each split, and repeating the first step until the records in each leaf node belong to the same class and grow to a complete tree.
First select bifurcation feature approach: based on the information gain of the information entropy, equation 1 represents the degree of confusion of the data, and may define the information entropy:
when there is only one data category in equation 1, i.e., p (x i )＝1，log 2 p(x i ) Since h=0, the purer the data, the more categories H is greater.
The information gain then needs to be defined as shown in equation 2. The information entropy of the child node is shown in formula 3.
Gain=h (D) -H' (D) formula 2
The difference between the information entropy of the parent node and the information entropy of the child node is called "category information entropy". The information entropy of the child node is called "condition information entropy" or "attribute information entropy". Information gain = category information entropy-attribute information entropy. When the information gain result is larger than and approaches to 1, the label is 1, and the index degradation is represented in the service, and vice versa.
Taking the operation log of 5g 2022, 10/13 in certain city as an example, the degradation problem of VoLTE service is determined. The model contains 35460 pieces of data, and the prediction accuracy of the model is 100%. The proving method can be used for processing corresponding service problems of the core network. As shown in table 1.
TABLE 1
The foregoing description is only illustrative of the preferred embodiments of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.
Claims (10)
1. A core network service quality problem identification method based on artificial intelligence is characterized in that,
and constructing a decision tree according to the related characteristic data in the service, modeling, judging the degradation of the service data through an optimization framework LGBM of the decision tree, and predicting the degradation of the subsequent service data.
2. The method of claim 1, wherein the step of determining the position of the substrate comprises,
firstly, data preprocessing:
modeling by taking the operation time as a demarcation point and adopting data with a time period of one month, wherein the granularity of the data is 5 minutes; and judging whether the service is deteriorated or not through the prediction of the data.
3. The method of claim 2, wherein the step of determining the position of the substrate comprises,
null value processing:
1.1 deleting characteristics containing missing values: if the deletion rate of the variable is more than 80%, directly deleting the variable;
1.2 mean value interpolation; the attribute of the data is divided into a fixed-distance type and a non-fixed-distance type, if the missing value is fixed-distance type, the missing value is interpolated by the average value of the existence value of the attribute; if the missing values are non-fixed, the missing values are complemented with the mode of the attribute (i.e., the value with the highest frequency of occurrence) according to the mode principle in statistics.
4. The method of claim 2, wherein the step of determining the position of the substrate comprises,
outlier processing:
absolute value difference median method: the method of detecting outliers by summing the distances between all factors and the average value is first required.
5. The method of claim 4, wherein the step of determining the position of the first electrode is performed,
logic for processing:
step one, finding out the median Xmedia of all factors;
step two, obtaining an absolute deviation value Xi-Xmedia of each factor and a median;
thirdly, obtaining a median MAD of the absolute deviation value; finally, the parameter n is determined, so that a reasonable range is determined as [ Xmedian-nMAD, xmedian+nmad ], and an adjustment is made for the factor values out of the reasonable range.
6. The method of claim 2, wherein the step of determining the position of the substrate comprises,
and the optimization framework LGBM based on the decision tree carries out identification prediction on the service problems.
7. The method of claim 6, wherein the step of providing the first layer comprises,
searching an attribute field with the maximum information quantity in a database by utilizing the information gain in the information theory, establishing a node of a decision tree, establishing branches of the tree according to different values of the attribute field, and repeatedly establishing lower nodes and branches of the tree in each branch subset.
8. The method of claim 7, wherein the step of determining the position of the probe is performed,
the concrete process of constructing the decision tree is as follows:
first, looking for an initial split, the entire training set is used as a set for generating a decision tree, and each record of the training set must be already classified to decide which attribute domain is the best classification index at present.
9. The method of claim 8, wherein the step of determining the position of the first electrode is performed,
firstly, exhausting all attribute domains, quantifying the quality of splitting of each attribute domain, and calculating the best splitting, wherein the quantification standard is to calculate the diversity index of each splitting;
next, the previous step is repeated until the records within each leaf node belong to the same class and grow to a complete tree.
10. The method of claim 9, wherein the step of determining the position of the substrate comprises,
first select bifurcation feature approach: based on the information gain of the information entropy, equation 1 represents the degree of confusion of the data, and may define the information entropy:
when there is only one data category in equation 1, i.e., p (x i )＝1，log 2 p(x i ) Because h=0, the purer the data, the more categories H is greater;
then the information gain needs to be defined as shown in equation 2; the information entropy of the child node is shown in a formula 3;
Gain＝H(D)-H′(D)
equation 2
The information entropy of the father node and the information entropy of the child node are different, and the information entropy of the father node is called as 'category information entropy'; the information entropy of the child node is called as "condition information entropy" or "attribute information entropy"; information gain = category information entropy-attribute information entropy; when the information gain result is larger than and approaches to 1, the label is 1, and the index degradation is represented in the service, and vice versa.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211545658.1A CN116016303A (en) | 2022-12-05 | 2022-12-05 | Method for identifying service quality problem of core network based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211545658.1A CN116016303A (en) | 2022-12-05 | 2022-12-05 | Method for identifying service quality problem of core network based on artificial intelligence |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116016303A true CN116016303A (en) | 2023-04-25 |
Family
ID=86030789
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211545658.1A Pending CN116016303A (en) | 2022-12-05 | 2022-12-05 | Method for identifying service quality problem of core network based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116016303A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117077043A (en) * | 2023-10-17 | 2023-11-17 | 深圳翱翔锐影科技有限公司 | Evaluation method of CdZnTe photon counting detector based on leakage current response |
| CN117075884A (en) * | 2023-10-13 | 2023-11-17 | 南京飓风引擎信息技术有限公司 | Digital processing system and method based on visual script |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104537010A (en) * | 2014-12-17 | 2015-04-22 | 温州大学 | Component classifying method based on net establishing software of decision tree |
| CN109446185A (en) * | 2018-08-29 | 2019-03-08 | 广西大学 | Collaborative filtering missing data processing method based on user's cluster |
| CN109711865A (en) * | 2018-12-07 | 2019-05-03 | 恒安嘉新（北京）科技股份公司 | A method of prediction is refined based on the mobile radio communication flow that user behavior excavates |
| CN110084424A (en) * | 2019-04-25 | 2019-08-02 | 国网浙江省电力有限公司 | A kind of Methods of electric load forecasting based on LSTM and LGBM |
| CN110837841A (en) * | 2018-08-17 | 2020-02-25 | 北京亿阳信通科技有限公司 | KPI (Key performance indicator) degradation root cause identification method and device based on random forest |
| CN112116014A (en) * | 2020-09-24 | 2020-12-22 | 贵州电网有限责任公司 | Test data outlier detection method for distribution automation equipment |
| WO2021240275A1 (en) * | 2020-05-08 | 2021-12-02 | Intime Biotech Llc | Real-time method of bio big data automatic collection for personalized lifespan prediction |
-
2022
- 2022-12-05 CN CN202211545658.1A patent/CN116016303A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104537010A (en) * | 2014-12-17 | 2015-04-22 | 温州大学 | Component classifying method based on net establishing software of decision tree |
| CN110837841A (en) * | 2018-08-17 | 2020-02-25 | 北京亿阳信通科技有限公司 | KPI (Key performance indicator) degradation root cause identification method and device based on random forest |
| CN109446185A (en) * | 2018-08-29 | 2019-03-08 | 广西大学 | Collaborative filtering missing data processing method based on user's cluster |
| CN109711865A (en) * | 2018-12-07 | 2019-05-03 | 恒安嘉新（北京）科技股份公司 | A method of prediction is refined based on the mobile radio communication flow that user behavior excavates |
| CN110084424A (en) * | 2019-04-25 | 2019-08-02 | 国网浙江省电力有限公司 | A kind of Methods of electric load forecasting based on LSTM and LGBM |
| WO2021240275A1 (en) * | 2020-05-08 | 2021-12-02 | Intime Biotech Llc | Real-time method of bio big data automatic collection for personalized lifespan prediction |
| CN112116014A (en) * | 2020-09-24 | 2020-12-22 | 贵州电网有限责任公司 | Test data outlier detection method for distribution automation equipment |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117075884A (en) * | 2023-10-13 | 2023-11-17 | 南京飓风引擎信息技术有限公司 | Digital processing system and method based on visual script |
| CN117075884B (en) * | 2023-10-13 | 2023-12-15 | 南京飓风引擎信息技术有限公司 | Digital processing system and method based on visual script |
| CN117077043A (en) * | 2023-10-17 | 2023-11-17 | 深圳翱翔锐影科技有限公司 | Evaluation method of CdZnTe photon counting detector based on leakage current response |
| CN117077043B (en) * | 2023-10-17 | 2024-01-30 | 深圳翱翔锐影科技有限公司 | Evaluation method of CdZnTe photon counting detector based on leakage current response |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116016303A (en) | Method for identifying service quality problem of core network based on artificial intelligence | |
| CN109685583B (en) | Supply chain demand prediction method based on big data | |
| De Winter et al. | Combining temporal aspects of dynamic networks with node2vec for a more efficient dynamic link prediction | |
| CN108154425B (en) | Offline merchant recommendation method combining social network and location | |
| Porteiro et al. | Electricity demand forecasting in industrial and residential facilities using ensemble machine learning | |
| CN110990718B (en) | Social network model building module of company image lifting system | |
| CN114221790A (en) | BGP (Border gateway protocol) anomaly detection method and system based on graph attention network | |
| CN115497272A (en) | Construction period intelligent early warning system and method based on digital construction | |
| CN112733996A (en) | GA-PSO (genetic Algorithm-particle swarm optimization) based hydrological time sequence prediction method for optimizing XGboost | |
| CN110008977B (en) | Clustering model construction method and device | |
| CN109787821B (en) | Intelligent prediction method for large-scale mobile client traffic consumption | |
| CN113723844B (en) | Low-voltage station theoretical line loss calculation method based on ensemble learning | |
| CN111210170A (en) | Environment-friendly management and control monitoring and evaluation method based on 90% electricity distribution characteristic index | |
| CN112149352B (en) | Prediction method for marketing activity clicking by combining GBDT automatic characteristic engineering | |
| CN111882157A (en) | Demand prediction method and system based on deep space-time neural network and computer readable storage medium | |
| CN106919564A (en) | A kind of influence power measure based on mobile subscriber's behavior | |
| CN116468536A (en) | Automatic risk control rule generation method | |
| CN115271041A (en) | Method for predicting telephone traffic of power service | |
| CN114372561A (en) | Network traffic prediction method based on depth state space model | |
| CN117453764A (en) | Data mining analysis method | |
| CN116415957A (en) | Abnormal transaction object identification method, device, computer equipment and storage medium | |
| Yu et al. | Forecasting digital economy of China using an Adaptive Lasso and grey model optimized by particle swarm optimization algorithm | |
| CN115809280A (en) | Group house renting identification and iteration identification method | |
| CN111343664A (en) | User positioning method, device, equipment and medium | |
| CN114820074A (en) | Target user group prediction model construction method based on machine learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |