In this section, we shall briefly discuss a few potential research and development solutions in order to address some of the issues mentioned:
- Distributed learning: In this book, the model learning or training were done centrally, but this may not be feasible in many IoT applications. In this context, distributed learning can be a potential solution. However, distributed computing has a security problem, and this can be minimized through Blockchain-based Distributed Learning.
- IoT mobile data: Smartphones are a key contributor of IoT proliferation. Efficient solutions for mobile big data analysis through DL models can offer better IoT services in various application domains, including smart cities. This area needs further research.
- Integration of contextual information: Contextual information is essential for correctly using and interpreting DL-based...
