Infusing Expert Knowledge into a Siamese Neural Network for Encoding Time Series
摘要
This chapter focuses on an application scenario that occurs later in the product life cycle, where cases of faults and failures are available to build a case-based reasoning (CBR) system. In this context, it presents the integration of expert knowledge into a Siamese neural network to serve as a similarity measure in the retrieval phase of the CBR system for implementing similarity-based predictive maintenance.