Abnormal user identification method and device based on XGBoost algorithm, and computer readable storage medium

Abstract

The invention relates to the field of artificial intelligence, and particularly discloses an abnormal user identification method and device based on an XGBoost algorithm and a computer readable storage medium, and the method comprises the implementation steps: data preprocessing and feature selection: obtaining user data needing to be identified in batches, and carrying out the data preprocessing through data cleaning and feature engineering; model establishing: constructing a classification model by taking the processed feature vector and the category label as a sample set input by the model, calculating a predicted value, then constructing an objective function of an algorithm according to the predicted value calculated and output by the model, and performing iteration to obtain an optimal loss function so as to obtain a final classification result; and model parameter tuning and model verification: optimizing the model parameters. According to the invention, by using the multi-dimensional user data, the coverage of the data is more comprehensive; a weak learner CART algorithm is selected, so that the operation efficiency is improved; the model expression effect is enhanced, the generalization ability is enhanced, the accuracy and the recognition rate are both improved, and computing resources are saved.

Description

technical field [0001] The present invention relates to the field of artificial intelligence, in particular to a method, system and equipment for identifying abnormal users based on the XGBoost algorithm. Background technique [0002] With the development of the big data era, the competition among operators in the communication industry is also increasing. In order to attract more users, operators carry out some marketing activities to attract new users who register from various channels. Some abnormally registered users have dealt with risky users such as not using mobile phone cards or maintaining cards, fraud, etc. In order to identify these abnormal users, operators use a large amount of data generated by users every day, such as user basic information, online behavior data, communication consumption data And the location data establishes an algorithm model to classify and identify abnormal users. [0003] At present, the identification of abnormal users in the communic...

AI Technical Summary

Problems solved by technology

[0011] In order to overcome the deficiencies of the prior art, the present invention provides an abnormal user identification method, device and computer-readable storage medium based on the XGBoost algorithm. Single, incomplete coverage problems; using integrated algorithms, integrating weak learners, improving model generalization ability and performance of model output, parallel optimization of feature granularity, improving algorithm efficiency and reducing calculation load

Images