Work

Hyper-parameter Optimization of Regularized-Extreme Learning Machine Using a Modified Opposition-Based Learning Artificial Bee Colony