Applicazioni di Deep Learning per Conditiopn Monitoring di componenti meccanici

Condition Monitoring is one of the crucial fields in the "industry 4.0" world. The aim of the thesis is developing robust Deep Learning Models for condition monitoring of mechanical components in a real working environments. Many different architectures, training approaches and models are exploited, studied and applied on the problem. The obtained results are compared with Machine Learning models created for solving the same task and developed by third part.