Main objective

Deploy the identified services in the co-creation labs and to connect them with the HPC infrastructure and with the applications, and the city services to be provided to users and stakeholders.

Tasks

• Task 2.1: Preparation: data models and analytical services
• Task 2.2: HPC and services set-up and deployment

Task 2.1 Preparation: data models and analytical services

Before the modelling phase for each case study, the data provided by the city of Lisbon was analysed to verify their consistency and quality to assure that the data provided has the adequate quality to be used in the modelling phase. This pre-processing focused in the existence and posterior imputation of missing values in the datasets, removal of duplicate values, uniformity of categorical variables.

After this process of data cleaning the datasets were aggregated and joined by space and time. More concretely a spatial grid was used for each case study to aggregate spatially the datasets. The spatial resolution and grid characteristics were dependent of the data characteristics and the question addressed in each case study. The same rational was used for the temporal aggregation, were the temporal aggregation periods were defined accordingly with the datasets characteristics along with the question addressed in each case study.

In the modelling phase are being used models that are generally used as solution for several problems, namely linear regression, support vector machines, random forests and neural networks.

Outputs:

• Data models and analytical services specifications (Micromobility; Waste Management; Parking; Pollution; Emergency).
• Services HPC configuration report (Final Version).
• Lisbon use cases implementation.
• Validation of datasets per service with the European Data Portal.
• Others (World Data League (WDL); Strategy and Insights Engine Project).