Telecommunication customer loss forecasting method based on nervous-netowrk improved algorithm

A neural network and improved algorithm technology, which is applied in the field of data mining of telecom operators, can solve the problems that the training error cannot be reduced, cannot be used to predict telecom customers, and the convergence speed is slow.

Inactive Publication Date: 2007-01-17
LINKAGE SYST INTEGRATION
View PDF1 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] 1. The problem of slow convergence speed limits the further application of this method
[0009] 2. The problem of local minima makes the training error unable to decrease and cannot be used to predict telecom customers (churn)

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Telecommunication customer loss forecasting method based on nervous-netowrk improved algorithm
  • Telecommunication customer loss forecasting method based on nervous-netowrk improved algorithm
  • Telecommunication customer loss forecasting method based on nervous-netowrk improved algorithm

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach

[0039] The first step: use the improved method to build a model,

[0040] The second step: data preparation. Due to the needs of the neural network model, the data is normalized. The normalization formula is: x ′ = 0.8 ( x - x min x max - x min ) + 0.1

[0041] Step 3: Train the model

[0042] After 231 times of training, the training error of the model is 0.000934405, which can meet the error accuracy requirements.

[0043] Not lost (account)

Lost (household)

actually

2628

195

predict

2125

698

predicted success

2109

179

...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The method comprises: firstly using BP neural network to build a client running off model; then using the error propagation factor to make normalization process for the data with a normalization formula: x'=0.8(x-xmin / xmax-xmin)+0.1, and wherein, x is an inputted parameter, xmin is a minimum inputted parameter, xmax is a maximum parameter, x is a normalized inputted parameter.

Description

technical field [0001] The invention belongs to the field of data mining of telecommunication operators, and relates to a method for predicting telecommunication customers (churn) based on an improved neural network algorithm. Background technique [0002] With the development of competition in the telecommunication market, customers have more and more room to choose telecommunication products and telecommunication companies, and the scramble for customers among telecommunication companies is becoming more and more fierce. Facing the increasingly fierce market competition environment, the traditional and passive service system of telecom companies can no longer meet the needs of customers and meet the challenges of competitors. At the same time, it is difficult for traditional advantages such as network and technology to widen the gap among telecom companies, and it is impossible to form a differentiated competitive advantage. Therefore, in order to c...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): H04L12/56H04L12/24H04L12/26H04W16/22
Inventor 黄晓颖薛庆童余志刚庄学阳李岩
Owner LINKAGE SYST INTEGRATION
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap