Efficient Hyper-parameter Optimization with Cubic Regularization

As hyper-parameters are ubiquitous and can significantly affect the model performance, hyper-parameter optimization is extremely important in machine learning.