Conclusion

Supervised statistical learning is a set of models and algorithms that has proven its usefulness and power in an ever-growing variety of application fields. It therefore has a rightful place (in the same way as the basics of signal processing, for example) in the "toolbox" of techniques and algorithms that every engineer should now be familiar with, if not at least master. All the more so as it is one of the keys to adding intelligence to products and services, which is one of the current challenges of innovation and international economic competition.

The field of statistical learning is currently booming, not only because of the proliferation of new applications, but also because of the vitality and productivity of research into the methods themselves: new models and algorithms are regularly invented and presented to the community. Deep learning methods are naturally...