业务推送及校准参数的获取方法和装置
By generating a calibration tree and dividing it into clusters through a two-stage calibration method, the problem of low calibration confidence caused by the sparsity of sample business data is solved, thereby improving the accuracy and confidence of business data push.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- RAJAX NETWORK &TECHNOLOGY (SHANGHAI) CO LTD
- Filing Date
- 2023-02-14
- Publication Date
- 2026-07-17
AI Technical Summary
During the business data push process, the sparse sample business data leads to low confidence in the calibration results, affecting the push effect of business data and making it difficult to migrate data from other business scenarios to the target business scenario to obtain calibration parameters.
A two-stage calibration method is adopted. First, a calibration tree is generated based on the first sample business data of multiple business scenarios. Then, the calibration tree is used to divide the second sample business data of the target business scenario into clusters to obtain calibration parameters, which solves the problem of sparse sample business data.
This improved the confidence level of calibration parameters, enhanced the delivery effect of business data, and ensured the accuracy and confidence level of calibration results.
Smart Images
Figure CN116342214B_ABST