Design of the 2015 ChaLearn AutoML challenge

摘要: ChaLearn is organizing the Automatic Machine Learning (AutoML) contest for IJCNN 2015, which challenges participants to solve classification and regression problems without any human intervention. Participants' code automatically run on servers train test learning machines. However, there no obligation submit code; half of prizes can be won by submitting prediction results only. Datasets progressively increasing difficulty are introduced throughout six rounds challenge. (Participants enter competition in round.) The alternate phases learners tested datasets have not seen, limited time tweak their algorithms those improve performance. This challenge will push state art fully automatic machine a wide range real-world problems. platform remain available beyond termination