CN101739614A - Hierarchy-combined prediction method for communication service - Google Patents
Hierarchy-combined prediction method for communication service Download PDFInfo
- Publication number
- CN101739614A CN101739614A CN200910232590A CN200910232590A CN101739614A CN 101739614 A CN101739614 A CN 101739614A CN 200910232590 A CN200910232590 A CN 200910232590A CN 200910232590 A CN200910232590 A CN 200910232590A CN 101739614 A CN101739614 A CN 101739614A
- Authority
- CN
- China
- Prior art keywords
- communication service
- prediction
- predicted value
- adaptability
- forecasting methodology
- 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
Images
Abstract
The invention discloses a hierarchy-combined prediction method for a communication service. The method comprises the the following steps of: using a communication service prediction as a target layer, wherein influence factors of the communication service prediction include fitting degree of the communication service prediction with historical data, compatibility of the communication service prediction with the future internal/external environmental change, adaptability of the communication service prediction to the policy of commun ication industry, adaptability of the communication service prediction to the competitive strategy of communication operation, adaptability of a prediction model of the communication service and the reliability of prediction model result; according to the influence factors, selecting n prediction methods with good fitting effect and high prediction precision, wherein the n prediction methods include a regression analysis method, a grey system model method, a trend extrapolation method, an growth curve method and a Rayleigh distribution multi-factor method; gaining n communication service prediction values by using the n prediction methods; and weighing or averaging the n communication service prediction values to obtain the communication service prediction value.
Description
Technical field
The present invention relates to communication service future development forecast method, the level built-up pattern is incorporated in the communication traffic prediction, designed one based on combination forecasting method, consider the Forecasting Methodology of the internal and external environment factor that influences the communication service development simultaneously, belong to the traffic forecast method of qualitative and quantitative combination.
Background technology
Telecommunication service prediction is the precondition of communication network stage construction plan, also is simultaneously telecommunication traffic and one of the necessary condition of estimate of income project period.The selection of Forecasting Methodology is directly connected to the realization of target of prediction and the levels of precision that predicts the outcome.A lot of about communications industry forecast method, common forecasting method is such as Forecasting Methodologies such as trend extrapolation, growth-curve approach, popularity rate method, the multifactor method of rayleigh distributed, regression analysis, Grey System Model predicted method, Bass modellings.Different Forecasting Methodologies extracts the information of sample data from different angles, at aspects such as complicacy, data demands difference is arranged all.In addition, general Forecasting Methodology is considered not enough to the factors such as confidence level of the fitting degree of economic implication, operator's service level and competitive strategy, Forecasting Methodology and historical datas such as the granting of the variation of communications industry policy such as 3G license, communications industry dissymmetry control policy, financial storm, Forecasting Methodology adaptability, Forecasting Methodology, therefore adopt general Forecasting Methodology to be difficult to obtain higher predicting the outcome of degree of accuracy.In the prediction of reality, can guarantee under any circumstance can both obtain to predict the outcome accurately without any a kind of method.
The present invention introduces step analysis thought on combined prediction theoretical foundation, the high factor of public trust of quantization influence prediction is set up the structural model of communication service prediction, thereby obtains the satisfactory solution and the recommendation of communication service prediction.
Summary of the invention
The objective of the invention is: the method that proposes a kind of hierarchy-combined prediction communication service, the high factor of public trust that the internal and external environment that taking into full account influences the communication service development changes, the key factor that the internal and external environment of quantization influence communication service development changes, a kind of Forecasting Methodology of raising precision of prediction.
Technical scheme of the present invention is: the method for hierarchy-combined prediction communication service, with the communication service predicted value as destination layer; The influence factor of communication service prediction is: with historical data fitting degree, the harmony with following internal and external environment variation, the adaptability with communications industry policy, the adaptability with the common carrier competitive strategy, the adaptability of communication service forecast model, forecast model result's confidence level; The factor according to above-mentioned influence is selected fitting effect and the higher n kind Forecasting Methodology of precision of prediction: n kind Forecasting Methodologies such as regression analysis, Grey System Model method, trend extrapolation, growth-curve approach, the multifactor method of rayleigh distributed; Above n kind Forecasting Methodology obtains n communication service predicted value; N communication service predicted value compared in twos, determine its relative significance level, and set up judgment matrix in view of the above, carry out consistance then and judge, revise, up to by consistency check for nonconforming square according to the 1-9 scaling law; Obtain relative weights; Obtain the communication service predicted value.
Perhaps n communication service predicted value averaged and obtain the communication service predicted value.
Perhaps n communication service predicted value carried out single preface, or carry out priority ordering.Determine the relative weights of each key element; N kind Forecasting Methodology obtains the communication service predicted value with the calculating and the normalizing of n communication service predicted value by comprehensive weights.
The beneficial effect of the inventive method: the method that a kind of new communication service that the prediction communication service result combinations of different Forecasting Methodology gained is got up to form predicts the outcome, this method has overcome the application category and the characteristics of various single Forecasting Methodologies self.The present invention has also overcome traditional Forecasting Methodology and has lacked the appropriate shortcoming of considering to influencing communications industry business development inside and outside factor.The high factor of public trust that influences the communication service prediction is carried out the mathematicization analysis, to a certain extent, improved the science and the validity of Forecasting Methodology, further improve the effect of communications industry traffic forecast.
Improve precision of prediction.Adopt the differential weights combination forecasting method, overcome different Forecasting Methodologies, improve precision of prediction in the influence of aspect such as complicacy, data line, sample size and method applicabilities to predicting the outcome.
Description of drawings
Fig. 1 communications industry combinations of services forecast level structural drawing.
Specific embodiments
Hierarchy-combined prediction has two kinds of citation forms: the one, and Deng power combination (method of average), the 2nd, differential weights combination (the present invention preferentially adopts, as 1-9 scaling law etc.).Suppose that in a certain forecasting problem the actual value of a certain period is y
t(t=1,2 ..., m), at this problem have the n kind mutually independently distinct methods wherein utilize i kind method that the predicted value of t period is y about the predicting the outcome of Y
It, the weight of establishing various Forecasting Methodologies is w=(w
1, w
2... w
n)
T. and satisfy w
1+ w
2+ ... + w
n=1,0＜w
i＜1 combination forecasting can be expressed as (t=1,2 ... m)
By proof, just like drawing a conclusion:
Suppose Y
1, Y
2Y
nBe the N kind mutually independently distinct methods is about the predicting the outcome of Y, this N kind Forecasting Methodology all is not have partially to estimate for Y.
Character 1: combined prediction Y also is not have partially to estimate for Y;
Character 2: combined prediction Y necessarily is better than single prediction.
2.2 Hierarchy Analysis Method
It is that a kind of multiple goal is estimated decision-making technique that the present invention adopts Hierarchy Analysis Method, it is decomposed into several levels and several key elements with the decision maker to the evaluation decision problem of complication system, between each key element, compare simply, judge and calculate, obtaining the weight of different key elements and different optional programs, thereby provide decision-making foundation for choosing the best alternatives.
The characteristics of analytical hierarchy process: with people's thought process mathematicization, systematization, so that accept.The needed quantitative information of this method seldom but requires the decision maker to grasp very thorough for the essence of decision problem, the key element that comprises and logical relation each other thereof.
The step of analytical hierarchy process:
(1) clear and definite problem is set up the multistage structural model of passing rank to the various key elements that constitute decision problem, promptly is destination layer (O)--rule layer ...--solution layer;
(2) be that criterion compares in twos to the key element of same grade (level) with the key element of upper level, determine its relative significance level according to the 1-9 scaling law, and set up judgment matrix in view of the above, carry out consistance then and judge, revise for nonconforming square, up to by consistency check;
(3) the single preface of level--by certain calculating, determine the relative weights of each key element;
(4) level is integrated ordered--by the calculating of comprehensive weights, various replacement schemes are carried out priority ordering.
2.3 hierarchy-combined prediction model
(1), adopt traffic forecast method as much as possible to predict according to communications industry business development historical data characteristics.Select fitting effect and the high n kind Forecasting Methodology of precision of prediction, obtain n kind prediction recommendation y
It(i=1,2...m).
(2) utilize Hierarchy Analysis Method to set up the professional level analytical structure of communications industry model.
Top (O): this one deck has only an element, generally is that it is the intended target or the desired result of problem analysis, therefore is called destination layer.We with the communication service predicted value as destination layer.
Rule layer: this one deck is to realize the related intermediate link of target.Communications industry traffic forecast mainly is subjected to the influence of following several respects factor: with historical data fitting degree, the harmony with following internal and external environment variation, the adaptability with communications industry policy, the adaptability with the common carrier competitive strategy, the adaptability of forecast model, forecast model result's confidence level etc.
Solution layer: this one deck is expressed as realizes the alternative various measures of target, decision scheme.Select fitting effect and the higher n kind Forecasting Methodology of precision of prediction, as solution layer.Such as: n kind Forecasting Methodologies such as forecasting by regression analysis, Grey System Model method, trend extrapolation, growth-curve approach, the multifactor method of rayleigh distributed.
Existing communication traffic forecast method has multiple: forecasting by regression analysis: forecasting by regression analysis is the mutual relationship that is presented according to two or more economic variable data (as per-capita gross domestic product, per capita consuming level, family's amount, resident population etc.), utilize historical data to set up regression equation (linear or non-linear, monobasic or polynary), find experimental formula specific between them, according to the variation of wherein one or more variablees, predict the development and change of another variable then.
The regression forecasting analysis is exactly by to one group of data analysis, sets up corresponding regression model, carries out parameter estimation, utilizes model the object of being studied is predicted and to be analyzed, and then provides foundation for economic decision-making.Regression Forecast has three major advantages: the one, can study the mutual relationship of forecasting object and correlative factor, and the essence reason of catching forecasting object to change, thereby it is more credible to predict the outcome; The 2nd, can provide the fiducial interval and the degree of confidence that predict the outcome, predict that complete sum is objective more thereby make; The 3rd, considered correlativity, can use relevant mathematical statistics method that regression equation is carried out statistical test, thereby the turning point that forecasting object is changed have certain distinguishing ability.
The Grey System Model method: grey forecasting model is called the GM model, is a kind of the system that contains uncertain factor to be carried out forecast method.Gray system is between unify system between the darky system of white color system.Gray system commonly used: grey systems GM (1,1) expression single order, the differential equation forecast model of a variable is mainly used in the seasonal effect in time series prediction.
X0: establishing time series X0 has n observed value X0={X1 (0), X2 (0), ... ..Xn (0) }, generate processing (Accumulaten Generating Operation is called for short AGO) through carrying out one-accumulate, generate new range X (1): X (1)={ X1 (1), X2 (1) ... ..Xn (1) }, wherein:
To this formation sequence, the differential equation of GM (1,1) model albefaction form is:
α is called the grey number of development in the formula; In being called, μ gives birth to the grey number of control.
If
Be parameter vector to be estimated,
Can utilize least square method to find the solution.Solve:
Find the solution the differential equation, get final product forecast model:
The characteristics of grey forecasting model: required sample size is few, and general sample size more than 4 is easy to use, does not need picture scatter diagram and repeatedly tentative calculation.Under the low sample size situation, precision of prediction is general.
Trend extrapolation: present certain rising or downtrending when forecasting object changed according to the time, not have tangible seasonal fluctuation, and can find a suitable function curve to reflect this variation tendency the time, just can adopt trend extrapolation to predict.It has reflected a kind of trend of market development, its certain reference value that predicts the outcome, but also there is certain limitation.Its prediction is to be based upon on the basic basis of invariable of market environment, is difficult to reflect the influence of following various variation to the market development rule, relatively is fit to last-period forecast.
Trend model comprises: polynomial curve Extrapolating model, index curve forecast model, logarithmic curve forecast model, growth curve trend extrapolation model, ridge amber be curve prediction model etc. now;
The multifactor method of rayleigh distributed: the multifactor method of rayleigh distributed is that forecast method is carried out in a kind of variation based on environmental baseline.The principle of this method is, meets the rule of rayleigh distributed substantially based on the distribution of people's income, the changes in permeability trend of research communication service in the potential user, thus obtain predicting the outcome to number of users.So-called potential user is meant the people of the PayPal communication service expense of having the ability, and promptly income surpasses the people of a certain threshold value.Because the difference of conditions of demand and consumer psychology has only the part potential user can use communication service to become the actual user.Communication service is subjected to influence of various factors such as price, rate, network service quality in the permeability of Potential Market, after these factors are quantized the influence of permeability, as long as can determine various influence factors situation of change in time span of forecast, just can quantitative forecast go out the permeability of the interior Potential Market of time span of forecast.
The concrete steps of the multifactor method of rayleigh distributed at first are the sizes of determining the potential market scale, then major influence factors are quantized, and determine the popularity rate P of communication service on potential market, and last two multiply each other and obtain predicted value.
Wherein: X
0Be thresholding (threshold value) that u is a per capita income
Wherein: P is the popularity rates of communication products on potential market
Δ f is the rate of change of various influence factors
R is weights
E is an elasticity coefficient
K is the number of influence factor
The method characteristics: based on the multifactor prediction of rayleigh distributed model, can better embody the substantial connection of economic development, the level of consumption and user development, be fit in the communication user, a kind of effective ways of long-term forecasting.
The Optimal Combination Forecasting method: combination forecasting method is a kind of brand-new Forecasting Methodology, and it is theoretical and method is progressively perfect, and range of application is constantly opened up extensively.Utilize combined prediction error sum of squares minimum, determine the method for optimal weighting coefficients to be called the Optimal Combination Forecasting method by the least square estimation method.If m kind forecast model is arranged, n period predicted for same forecasting problem.Note: y
tBe actual value, y
t *Be with the predicted value of combination forecasting method to the t period; y
It *Be the predicted value of i kind Forecasting Methodology to the t period; e
ItBe the predicated error of i kind Forecasting Methodology to the t period; w
iBe the flexible strategy of i kind Forecasting Methodology, satisfy
e
iFor with the predicated error of combination forecasting method to the t period.I=1 wherein, 2,3..., m; T=1,2...n. then can obtain:
e
It=y
t-y
It *, the Optimal Combination Forecasting method is determined that the problem of weight is converted to and is asked optimization problem: in constraint condition
Have
Find the solution above-mentioned optimization problem by the least square estimation method, obtain the optimal weights value.With reference to figure 1 communications industry combined prediction hierarchical chart.
The model of the communications industry combinations of services prediction of more than setting up is the decision model that takes into full account the various inside and outsides factor that influences communication service prediction recommendation, carries out mathematical computations by EXCEL, thereby obtains optimum weight w
i(i=1,2 ... m).
(3) utilize differential weights combined prediction method, obtain the optimum recommendation of communication service prediction, promptly be,
With 2005-2007 number of mobile users of certain districts and cities is sample, and we adopted step analysis combined prediction method to predict certain districts and cities mobile subscriber development below sample data saw table 1. for details.
2005-2007 number of mobile users of certain districts and cities of table 1
| Time | ??2005.01 | ??2005.02 | ??2005.03 | ??2005.04 | ??2005.05 | ??2005.06 |
| Number of mobile users | ??340.4 | ??351.8 | ??368.7 | ??372.5 | ??381.2 | ??381.0 |
| Time | ??2005.07 | ??2005.08 | ??2005.09 | ??2005.10 | ??2005.11 | ??2005.12 |
| Number of mobile users | ??384.3 | ??389.0 | ??397.7 | ??413.2 | ??419.7 | ??423.5 |
| Time | ??2006.01 | ??2006.02 | ??2006.03 | ??2006.04 | ??2006.05 | ??2006.06 |
| Number of mobile users | ??436.6 | ??456.9 | ??475.4 | ??480.0 | ??493.0 | ??495.0 |
| Time | ??2006.07 | ??2006.08 | ??2006.09 | ??2006.10 | ??2006.11 | ??2006.12 |
| Number of mobile users | ??503.7 | ??513.5 | ??523.5 | ??545.9 | ??552.5 | ??559.5 |
| Time | ??2007.01 | ??2007.02 | ??2007.03 | ??2007.04 | ??2007.05 | ??2007.06 |
| Number of mobile users | ??564.2 | ??574.5 | ??621.2 | ??631.6 | ??648.8 | ??655.2 |
| Time | ??2007.07 | ??2007.08 | ??2007.09 | ??2007.10 | ??2007.11 | ??2007.12 |
| Number of mobile users | ??667.2 | ??677.2 | ??688.9 | ??703.6 | ??713.3 | ??722.9 |
1. single Forecasting Methodology prediction
Adopt the multiple business Forecasting Methodology certain districts and cities mobile subscriber to be predicted, choose 4 kinds of single preferably Forecasting Methodologies of prediction effect based on above-mentioned historical data.Its recommendation sees Table 2.
The single Forecasting Methodology of table 2 predicts the outcome to mobile service
| Forecasting Methodology | ??2008.01 | ??2008.02 | ??2008.03 | ??2008.04 | ??2008.05 | ??2008.06 |
| Trend extrapolation | ??738.52 | ??752.11 | ??765.76 | ??779.47 | ??793.24 | ??807.05 |
| The popularity rate pairing comparision | ??736.79 | ??759.48 | ??772.52 | ??785.56 | ??798.60 | ??811.65 |
| Growth-curve approach Forecasting Methodology trend extrapolation popularity rate pairing comparision growth-curve approach | ??752.00??2008.07??820.91??824.69??857.64 | ??768.66??2008.08??834.81??837.73??876.63 | ??785.69??2008.09??848.73??850.77??896.05 | ??803.09??2008.10??862.68??863.81??915.90 | ??820.88??2008.11??876.66??876.85??936.19 | ??839.06??2008.12??890.65??889.90??956.92 |
2. utilize Hierarchy Analysis Method, make up mobile subscriber's step analysis structural model.
(1) uses each layer of 1-9 gradation calculations judgment matrix.
After stratum's aggregated(particle) structure was passed in foundation, the membership of element was determined between levels was inferior.With certain factor of the relative last layer of the key element of same level in the hierarchy Model time, be carried out to right comparison each other and form the comparison judgment matrix.Each element of rule layer is for the positive Reciprocal Judgement Matrix of prediction of overall target, and is as shown in table 3.
Each factor of table 3 rule layer is for the comparison judgment matrix of prediction general objective
| Number of mobile users predicts the outcome | The historical data fitting degree | With future economy harmony | Forecast model adaptability | Credibility predicts the outcome |
| The historical data fitting degree | ??1.00 | ??3.00 | ??5.00 | ??7.00 |
| With future economy harmony | ??0.33 | ??1.00 | ??5.00 | ??3.00 |
| The adaptability of model | ??0.20 | ??0.20 | ??1.00 | ??1.00 |
| The credibility that predicts the outcome | ??0.14 | ??0.33 | ??1.00 | ??1.00 |
Maximum proper vector: (0.56,0.27,0.08,0.08), λ
Max=4.11.
Consistance is judged: coincident indicator C.I=0.04, mean random coincident indicator rate C.R=0.039＜0.1.
(2) in like manner construct the comparison judgment matrix of Forecasting Methodology with respect to each rule layer, and carry out consistance and judge, find the solution the maximum characteristic root of judgment matrix, maximum proper vector, judgment matrix is as shown in table 4.
Table 4 Forecasting Methodology is with respect to the comparison judgment matrix of rule layer (historical data)
| The historical data fitting degree | Trend extrapolation | The popularity rate pairing comparision | Growth-curve approach |
| Trend extrapolation | ??1.00 | ??3.00 | ??9.00 |
| The popularity rate pairing comparision | ??0.33 | ??1.00 | ??2.00 |
| Growth-curve approach | ??0.11 | ??0.50 | ??1.00 |
Maximum proper vector: (0.7,0.21,0.09), λ
Max=3.02.
Consistance is judged: coincident indicator C.I=-0.33, mean random coincident indicator rate C.R=-0.36＜0.1.
The same can the structure forecast method with respect to the comparison judgment matrix of other rule layers, and carry out that consistance is judged and the maximum characteristic root of judgment matrix, the calculating of maximum proper vector.
(3) level is integrated ordered
Integrated ordered through level, we obtain the ordering of Forecasting Methodology weights, and are as shown in table 5.
Table 5 Forecasting Methodology weights
| Forecasting Methodology | Weights |
| Trend extrapolation | ??0.68 |
| The popularity rate pairing comparision | ??0.22 |
| Growth-curve approach | ??0.09 |
3. hierarchy-combined prediction model result
Certain districts and cities' communication service predicts the outcome, and is as shown in table 6.
Table 6 communication service step analysis combined prediction result
| ??2008.0 | ??2008.0 | ??2008.0 | ??2008.08 | ??2008.10 | ??2008.1 | Weights | |
| Trend extrapolation | ??738.52 | ??779.47 | ??807.05 | ??834.81 | ??862.68 | ??890.65 | ??0.68 |
| The popularity rate pairing comparision | ??736.79 | ??785.56 | ??811.65 | ??837.73 | ??863.81 | ??889.90 | ??0.22 |
| Growth-curve approach | ??752 | ??803.09 | ??839.06 | ??876.63 | ??915.90 | ??956.92 | ??0.09 |
| The hierarchy-combined prediction model | ??731.97 | ??775.14 | ??802.87 | ??830.66 | ??859.09 | ??887.54 | |
| Real data | ??713.49 | ??785.49 | ??802.5 | ??827.07 | ??848.24 | ??880.18 |
| ??2008.0 | ??2008.0 | ??2008.0 | ??2008.08 | ??2008.10 | ??2008.1 | Weights | |
| Error rate | ??-2.59％ | ??1.32％ | ??-0.05％ | ??-0.46％ | ??-1.28％ | ??-0.84％ |
Find relatively that by step analysis combined prediction result and historical data maximum error rate is-2.59%, the least error rate is-0.05%, and the average error rate that predicts the outcome has improved the precision of recommendation well below single Forecasting Methodology error rate.
The present invention proposes communication service Forecasting Methodology, set up the model structure of step analysis and made up its positive Reciprocal Judgement Matrix, by the weight of fixing excel modelling calculation combination Forecasting Methodology based on the combined prediction theory.Example shows that model method calculates all has certain advantage, can obtain the satisfactory solution of prediction, is science in the communications industry traffic forecast, effective, sufficient model and method for solving.
Claims (3)
1. the method for hierarchy-combined prediction communication service, it is characterized in that with the communication service predicted value that as destination layer the influence factor of communication service prediction is: the harmony that changes with the historical data fitting degree, with following internal and external environment, adaptability, adaptability, the adaptability of communication service forecast model, forecast model result's confidence level with the common carrier competitive strategy with communications industry policy; The factor according to above-mentioned influence is selected fitting effect and the higher n kind Forecasting Methodology of precision of prediction: n kind Forecasting Methodologies such as regression analysis, Grey System Model method, trend extrapolation, growth-curve approach, the multifactor method of rayleigh distributed; Above n kind Forecasting Methodology obtains n communication service predicted value; N communication service predicted value compared in twos, determine its relative significance level, and set up judgment matrix in view of the above, carry out consistance then and judge, revise, up to by consistency check for nonconforming square according to the 1-9 scaling law; Obtain relative weights; Thereby obtain the communication service predicted value.
2. the method for hierarchy-combined prediction communication service, it is characterized in that with the communication service predicted value that as destination layer the influence factor of communication service prediction is: the harmony that changes with the historical data fitting degree, with following internal and external environment, adaptability, adaptability, the adaptability of communication service forecast model, forecast model result's confidence level with the common carrier competitive strategy with communications industry policy; The factor according to above-mentioned influence is selected fitting effect and the higher n kind Forecasting Methodology of precision of prediction: n kind Forecasting Methodologies such as regression analysis, Grey System Model method, trend extrapolation, growth-curve approach, the multifactor method of rayleigh distributed; Above n kind Forecasting Methodology obtains n communication service predicted value; N communication service predicted value carried out single preface or carried out priority ordering.Determine the relative weights of each key element; N kind Forecasting Methodology obtains the communication service predicted value with the calculating and the normalizing of n communication service predicted value by comprehensive weights.
3. the method for hierarchy-combined prediction communication service, it is characterized in that with the communication service predicted value that as destination layer the influence factor of communication service prediction is: the harmony that changes with the historical data fitting degree, with following internal and external environment, adaptability, adaptability, the adaptability of communication service forecast model, forecast model result's confidence level with the common carrier competitive strategy with communications industry policy; The factor according to above-mentioned influence is selected fitting effect and the higher n kind Forecasting Methodology of precision of prediction: n kind Forecasting Methodologies such as regression analysis, Grey System Model method, trend extrapolation, growth-curve approach, the multifactor method of rayleigh distributed; Above n kind Forecasting Methodology obtains n communication service predicted value; N communication service predicted value averaged obtain the communication service predicted value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910232590A CN101739614A (en) | 2009-12-08 | 2009-12-08 | Hierarchy-combined prediction method for communication service |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910232590A CN101739614A (en) | 2009-12-08 | 2009-12-08 | Hierarchy-combined prediction method for communication service |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101739614A true CN101739614A (en) | 2010-06-16 |
Family
ID=42463070
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910232590A Pending CN101739614A (en) | 2009-12-08 | 2009-12-08 | Hierarchy-combined prediction method for communication service |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101739614A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017161646A1 (en) * | 2016-03-23 | 2017-09-28 | 南京华苏科技有限公司 | Method for dynamically selecting optimal model by three-layer association for large data volume prediction |
| CN107884708A (en) * | 2017-10-18 | 2018-04-06 | 广东电网有限责任公司佛山供电局 | A kind of switch performance diagnostic method based on switch service data |
| CN108268967A (en) * | 2017-01-04 | 2018-07-10 | 北京京东尚科信息技术有限公司 | A kind of method and system of traffic forecast |
| CN108874748A (en) * | 2018-06-01 | 2018-11-23 | 交通运输部路网监测与应急处置中心 | A kind of ETC operation indicator prediction technique |
| CN109034905A (en) * | 2018-08-03 | 2018-12-18 | 四川长虹电器股份有限公司 | The method for promoting sales volume prediction result robustness |
| CN109756911A (en) * | 2019-01-31 | 2019-05-14 | 腾讯科技（深圳）有限公司 | Network quality prediction technique, business reorganization method, relevant device and storage medium |
| CN111045907A (en) * | 2019-12-12 | 2020-04-21 | 苏州博纳讯动软件有限公司 | System capacity prediction method based on traffic |
| CN116596169A (en) * | 2023-07-17 | 2023-08-15 | 国网浙江省电力有限公司宁波供电公司 | Power system prediction method, device and storage medium |
-
2009
- 2009-12-08 CN CN200910232590A patent/CN101739614A/en active Pending
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017161646A1 (en) * | 2016-03-23 | 2017-09-28 | 南京华苏科技有限公司 | Method for dynamically selecting optimal model by three-layer association for large data volume prediction |
| CN108268967A (en) * | 2017-01-04 | 2018-07-10 | 北京京东尚科信息技术有限公司 | A kind of method and system of traffic forecast |
| CN107884708A (en) * | 2017-10-18 | 2018-04-06 | 广东电网有限责任公司佛山供电局 | A kind of switch performance diagnostic method based on switch service data |
| CN108874748A (en) * | 2018-06-01 | 2018-11-23 | 交通运输部路网监测与应急处置中心 | A kind of ETC operation indicator prediction technique |
| CN109034905A (en) * | 2018-08-03 | 2018-12-18 | 四川长虹电器股份有限公司 | The method for promoting sales volume prediction result robustness |
| CN109756911A (en) * | 2019-01-31 | 2019-05-14 | 腾讯科技（深圳）有限公司 | Network quality prediction technique, business reorganization method, relevant device and storage medium |
| CN109756911B (en) * | 2019-01-31 | 2020-11-27 | 腾讯科技（深圳）有限公司 | Network quality prediction method, service adjustment method, related device and storage medium |
| CN111045907A (en) * | 2019-12-12 | 2020-04-21 | 苏州博纳讯动软件有限公司 | System capacity prediction method based on traffic |
| CN116596169A (en) * | 2023-07-17 | 2023-08-15 | 国网浙江省电力有限公司宁波供电公司 | Power system prediction method, device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101739614A (en) | Hierarchy-combined prediction method for communication service | |
| Tatari et al. | Cost premium prediction of certified green buildings: A neural network approach | |
| CN103024762B (en) | Service feature based communication service forecasting method | |
| Bakhtiari et al. | A coupled agent-based risk-based optimization model for integrated urban water management | |
| Li et al. | Urban water demand forecasting based on HP filter and fuzzy neural network | |
| Lu et al. | Urban water security evaluation based on similarity measure model of Vague sets | |
| Kang et al. | Multicriteria Evaluation of Water Resources Sustainability in the Context of Watershed Management 1 | |
| CN104504280A (en) | Planning-demand-considered comprehensive evaluation method for communication mode of cluster management system of charging piles of electric automobile | |
| CN110490409B (en) | DNN-based low-voltage transformer area line loss rate benchmarking value setting method | |
| CN107292534A (en) | The yardstick competition evaluation method and device of urban power distribution network long term dynamics investment | |
| CN109919356A (en) | One kind being based on BP neural network section water demand prediction method | |
| CN110428270A (en) | The potential preference client recognition methods of the channel of logic-based regression algorithm | |
| CN111008870A (en) | Regional logistics demand prediction method based on PCA-BP neural network model | |
| CN106327099A (en) | Communication network comprehensive performance evaluation parameter weight determination and adjusting method | |
| CN105224997A (en) | A kind of planning load forecasting method containing multiple uncertain factor impact | |
| Caceres et al. | Estimating traffic flow profiles according to a relative attractiveness factor | |
| Xu et al. | A multi-object decision-making method for location model of manufacturing industry under uncertain environment | |
| CN108805471A (en) | Evaluation method for water resources carrying capacity based on the analysis of hybrid system interactively | |
| CN109933901B (en) | MCR city expansion simulation method for optimizing service value of ecosystem | |
| CN111507507A (en) | Big data-based monthly water consumption prediction method | |
| CN104866930A (en) | Time series prediction method based on quantitative screening time series prediction model | |
| CN102708298B (en) | A kind of Vehicular communication system electromagnetic compatibility index distribution method | |
| García-López et al. | Income segregation and urban spatial structure: Evidence from Brazil | |
| CN105160448A (en) | Electricity transaction monitoring risk index evaluation method based on unascertained rational numbers | |
| Kalan | The relationship between economic growth and road infrastructure: Evidence from Turkey |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20100616 |