Training analysis of optimization models in machine learning

Machine learning is fast evolving, with numerous theoretical advances and applications in a variety of domains. In reality, most machine learning algorithms are based on optimization issues. This interaction is also explored in the special topic on machine learning and large-scale optimization. Furthermore, machine learning optimization issues have several unique characteristics that are rarely seen in other optimization contexts. Aside from that, the notions of classical optimization vs machine learning will be discussed. Finally, this study will give an outline of these peculiar aspects of machine learning optimization.