Automated Hyperparameter Optimization Challenge at CIKM 2021 AnalyticCup

TL;DR

This paper presents a Bayesian optimization approach using the OpenBox package to automate hyperparameter tuning, achieving first place in the CIKM 2021 AnalyticCup competition with high accuracy results.

Contribution

The authors introduce a successful hyperparameter optimization method combining Bayesian optimization and early stopping, applied to industrial tasks in a competitive setting.

Findings

Achieved top performance in the CIKM 2021 challenge

Demonstrated effectiveness of Bayesian optimization with OpenBox

Secured first place in both preliminary and final contests

Abstract

In this paper, we describe our method for tackling the automated hyperparameter optimization challenge in QQ Browser 2021 AI Algorithm Competiton (ACM CIKM 2021 AnalyticCup Track 2). The competition organizers provide anonymized realistic industrial tasks and datasets for black-box optimization. Based on our open-sourced package OpenBox, we adopt the Bayesian optimization framework for configuration sampling and a heuristic early stopping strategy. We won first place in both the preliminary and final contests with the results of 0.938291 and 0.918753, respectively.