Читать книгу Digital Cities Roadmap - Группа авторов - Страница 48

The key ML activities that are applicable to SB will be identified. For the general description of ML activities in SBs and measures to incorporate ML in an SB setting the reader is alluded to in Figure 1.24.

Collecting and collecting data: A range of methods were used to collect data, each of varying resources, energy consumption and networking deals. Sensors and related artifacts in SBs simultaneously produce raw information and these devices can store or record the information on monitored components for a specified period of time.

Figure 1.24 ML tasks in SB Environment.

It can be utilized by decision-makers, planners, running and sustaining staff and building customers, many healthcare services and so on.

Data Pre-processing: Much data is generated in SBs by sensors from various sources with specific formats and architectures. The data come from different sources. This knowledge is not usually ready to be evaluated, since its poor battery capacity, bad tuning, access to numerous harmful elements and intervention may be incomplete or redundant.

Dimensionality Reduction: Raising volumes of raw data from heterogenous and all-embracing sensors used in SBs are enormous. The bulk of data from these sensors is redundant and needs to be minimized by utilizing techniques to limit their dimensionality to a smaller number of features without missing any valuable details.