COVID-19 battle

To understand the past development of Covid-19 cases and predict the development of the number cases in the future we need good source data, analytical models and the right visualization tools.

The quality of the underlying data is crucial as it determines the possibilities for the statistical analysis as well as for the visualization of the output of the prediction models.

We implemented this project together with Graphileon, our long-term partner and member of the CovidGraph initiative, a non-profit collaboration of researchers, software developers, data scientists and medical professionals.

Using our melda.io platform we implemented a statistical model (based upon an existing open source SIR model *1) to predict the development of case numbers for each country, as well as the dates by which 95% and 98% of all COVID-19 cases should have occurred. Graphileon’s graph-based platform was used as an interface to CovidGraph’s knowledge graph. Subsequently, the data was sent to our melda.io data-analytics service.
The result were then sent back to Graphileon, in order to be visualized in a dashboard.

*1-Statistical analysis on melda.io

Marisa Eisenberg’s SIR epidemic spread of disease model implementation (https://github.com/epimath/param-estimation-SIR(MIT License Copyright)) formed the basis for our analysis. We used this model to calculate the daily numbers of newly infected people which were based on the daily COVID-19 new case data per country from the Our World in Data COVID-19 dataset (https://ourworldindata.org/coronavirus-source-data). Next, we predicted the total expected number of cases for each country and used the total expected case numbers to predict the dates by which 95% and 98% of all COVID-19 cases should have occurred. These dates provide a valuable insight regarding the expected rate at which the pandemic spreads and allow decision-makers to make data-driven decisions about the loosening or tightening of restrictive measures. To view the open source implementation, go to : https://app.melda.io/view/graphileon-info/covid-end-date-prediction/v2
(a free account can be obtained by registering)

About melda.io

melda.io, a Modern, ELegant Data Analysis platform is available as a Software As A Service (SAAS) platform. It combines a scalable multi-language computation engine with a modern Integrated Development Environment (IDE). Melda.io provides the user with a notebook interface for coding and data visualization, an intelligent Data Science knowledgebase and multi-user collaboration. The melda platform was designed to make it easy for Data Scientists to manage code and data while co-development and interaction between recognized expert individuals and institutes in the Data Science industry is facilitated. A freely accessible Community Edition allow data scientists to use the platform for personal projects. More information can be found on www.melda.io

About Graphileon

Graphileon (www.graphileon.com) provides a unique graph-based application building platform, allowing non-developers to configure browser-based applications. Its products provide business consultants and information analysts with tools to rapidly design and deploy graph-based applications by exploiting the agility of graphs development. Graphileon has built-in support for multiple graph databases.