A method for predicting the risk of power outage complaints based on gradient boosting trees

Abstract

The invention relates to a method for predicting the risk of power outage complaints based on a gradient boosting tree, comprising the following steps: step A: establishing a user power consumption information table; step B: performing preprocessing on the user power consumption information data set in the user power consumption information table ; Step C: Use Canopy algorithm, KMeans The algorithm clusters the data sets of user electricity consumption information, and marks the sensitive categories of the data sets of user electricity consumption information through customer portrait analysis, and then uses SPARK-based SMOTE The oversampling algorithm performs data processing on the unbalanced distribution of user electricity information data sets; step D: conduct gradient boosting tree training on user electricity information data sets to obtain a power outage complaint risk model; step E: use the power outage complaint risk model, Predict outage sensitive categories of users. This method is beneficial to accurately determine the sensitivity of different users to power outages, so as to adopt different comfort and guidance strategies to reduce the number of users' complaints about power outages.

Description

technical field [0001] The invention relates to the technical field of power failure complaint risk prediction, in particular to a method for power failure complaint risk prediction based on Gradient-boosted trees. Background technique [0002] At present, all kinds of customers, including enterprises and individuals, have continuously improved expectations for the quality of power supply services, and put forward higher requirements for service quality. However, in the existing technology, power supply enterprises are mostly unable to classify the power outage sensitive types of different users, so as to adopt different appeasement and guidance strategies according to the sensitivity of different users to power outages, thus bringing a large number of power outage complaints and causing damage to the corporate image. The negative impact will cause troubles to the normal operation of the enterprise, and even lead to various legal disputes. Contents of the invention [000...

AI Technical Summary

Problems solved by technology

However, in the existing technology, power supply enterprises are mostly unable to classify the power outage sensitive types of different users, so as to adopt different appeasement and guidance strategies according to the sensitivity of different users to power outages, thus bringing a large number of power outage complaints and causing damage to the corporate image. Negative impact, causing troubles to the normal operation of the enterprise, and even deriving various legal disputes

Images