报告题目：Auto-WEKA Combined Selection and Hyperparameter Optimization of Classification Algorithms

报 告 人：梁青青

报告时间：2017年11月22日（周三） 下午 2:00

报告地点：贵州大学北校区博学楼603室

报告内容摘要：

This paper Many different machine learning algorithms exist; takinginto account each algorithm’s hyperparameters, there is astaggeringly large number of possible alternatives overall. Weconsider the problem of simultaneously selecting a learningalgorithm and setting its hyperparameters, going beyondprevious work that attacks these issues separately. We showthat this problem can be addressed by a fully automatedapproach, leveraging recent innovations in Bayesian optimization.