Method of evaluating advertisement conversion rate prediction models and server

Abstract

The embodiment of the invention discloses a method of evaluating advertisement conversion rate prediction models, servers and a computer-readable storage medium. The method includes: receiving an advertisement conversion rate prediction model calling request, randomly calling a new-version advertisement conversion rate prediction model or an old-version advertisement conversion rate prediction model to calculate predicted conversion rates of candidate advertisements, and recording model calling identifiers corresponding to the candidate advertisements; sending the predicted conversion rates ofthe candidate advertisements to an advertisement platform; calculating first actual advertisement conversion rates, which corresponds to the new-version advertisement conversion rate prediction model, and second actual advertisement conversion rates, which corresponds to the old-version advertisement conversion rate prediction model, according to model calling identifiers corresponding to alreadyexposed advertisements; and carrying out advantage / disadvantage evaluation on the new-version advertisement conversion rate prediction model and the old-version advertisement conversion rate prediction model according to the first actual advertisement conversion rates, the second actual advertisement conversion rates and a preset evaluation strategy. The embodiment of the invention can improve accuracy of advantage / disadvantage evaluation of the advertisement conversion rate prediction models.

Description

technical field [0001] The present invention relates to the field of communication technologies, in particular to a method for evaluating an advertisement conversion rate prediction model, a server and a computer-readable storage medium. Background technique [0002] The advertising conversion rate estimation model is used by the advertising platform to estimate the conversion rate (such as click rate or download rate) of the advertisement to be displayed before displaying the advertisement. When the version of the advertising conversion rate estimation model is updated, it is usually necessary to test and evaluate the new version of the advertising conversion rate estimation model to determine whether it is really better than the old version of the advertising conversion rate estimation model, and then determine whether it is based on the evaluation results. Launched a new version of the advertising conversion rate prediction model. [0003] The existing advertising conver...

AI Technical Summary

Problems solved by technology

[0003] The existing advertising conversion rate estimation model evaluation methods usually evaluate the new and old versions of the advertising conversion rate estimation model based on the real-time data of the new version of the advertising conversion rate estimation model and the historical data of the old version of the advertising conversion rate estimation model, that is, according to Evaluate the advertising conversion rate data estimated by the old version of the advertising conversion rate estimation model in the process of using it and the real-time estimated advertising conversion rate data of the new version of the advertising conversion rate estimation model during the test phase, that is, the existing The evaluation method of the advertising conversion rate prediction model is based on the data of different time dimensions when evaluating the old and new versions of the advertising conversion rate prediction model, which will lead to differences between the evaluation results and the actual advantages and disadvantages of the old and new versions. inaccurate result

Images