A data processing method, device, and computer-readable storage medium

Abstract

The embodiment of the present application discloses a data processing method, device, and computer-readable storage medium, wherein the method includes: in the Nth round of comparison test, acquiring the test duration and sampling index values ​​corresponding to at least two test objects respectively; when there is When the sampling index value of the object update condition is satisfied, the object to be updated is determined in the dynamic test object according to the test duration; the object optimization algorithm is determined according to the sampling index value of the fixed test object in each round of comparison test; Update the object, obtain the object to be tested in the test object set; update the object to be updated in the at least two test objects to the object to be tested, obtain at least two updated test objects, and determine the updated at least two test objects as Test subjects in round N+1 of controlled testing. By adopting the application, the amount of data calculation can be saved, and the cost of space calculation can be reduced.

Description

technical field [0001] The present application relates to the technical field of the Internet, and in particular to a data processing method, device, and computer-readable storage medium. Background technique [0002] In order to achieve continuous development, enterprises need to continuously make effective changes to the landing products, but it is also a very risky thing to change the perfect landing products at will, so many enterprises will make decisions through AB testing. [0003] The existing AB test is to conduct experiments on the control version and at least two test versions at the same time, wherein the control version is an existing product, and the test version is a version with changed version parameters, which is used to update the control version; then observe user business in real time Data, when the experiment duration is equal to the set experiment stop duration, the experiment is stopped at the same time for the control version and at least two test ve...

AI Technical Summary

Problems solved by technology

Obviously, the existing AB test has the following defects: 1) If there are a large number of test parameters, such as the color of a certain button, which can be set to hundreds or thousands of colors, there will be hundreds or thousands of test versions. Conduct comparative tests on hundreds or thousands of test versions at the same time, which involves a large amount of data calculations and has extremely high requirements for space computing costs; 2) The actual experiment duration is equal to the fixed experiment stop duration, and the existing AB test only focuses on The final result data of the experiment, while ignoring the data during the experiment, makes the test subjects with obviously poor experimental results continue the experiment meaninglessly until the experiment duration is equal to the experiment stop duration, which will also cause a waste of computing resources

Images