Algorithms are not innocent codes based on machine learning and artificial intelligence, but ideological devices that shape ...

Several sampling-based algorithms have been recently proposed that ensure asymptotic optimality. The convergence of these algorithms can be improved if sampling is guided toward the most promising ...

Multiclass classification is one of the most fundamental tasks in data mining. However, traditional data mining methods rely on the model assumption, they generally can suffer from the overfitting ...