Biometric Authentication: A Machine Learning Approach

6.4. Hierarchical Machine Learning Models

The information made available by the functional modules in a modular network must be properly integrated to reach a final decision or recommendation. There are many possible techniques to integrate or fuse information from various local modules. The most popular approach uses some kind of hierarchical structure. For example, under a hierarchical model, a set of class-based modules can collectively form an expert subnet in expert-based type networks. Therefore, both the notions of expert modules and class modules play a vital role in the construction of such a hierarchical learning network.

According to the levels of hierarchy involved, hierarchical networks can be divided structurally into the following ...