top of page

Covid19 Demo

The data presented in herby is based on the data from "COVID-19 Data Repository" by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University: The data concerning country names and population size are taken from the World Bank:

Yokaranda does not verify or guarantee the correctness of the data or of its processing. The results presented are intended only to demonstrate automatic ML capabilities and should not be used for any other purposes

Population Type

Making the Correct Decisions to Contain COVID-19

Yokaranda makes AI/ML Application monitoring simple and powerful and helps developers ensure the successful and reliable rollout of their business-critical AI/ML applications. Its powerful monitoring libraries and SDK are used to analyze large and diverse data sets from IoT devices, mobile phones, and other data sources. In this case, we would like to illustrate the benefits of Yokaranda's monitoring solution in analyzing international COVID-19 public data sources.

The Challenge

Governments are overwhelmed with a massive amount of COVID-19 data and are challenged to make the correct lockdown restrictions to balance public health and economic constraints. While COVID-19 data is piling up, it is challenging to gain valuable insights unless you use a powerful AI/ML solution. 

The Solution

The application uses the Yokaranda libraries to load publicly available COVID-19 data sources and break them into "time series" of COVID-19 data across time. It learns the behavior of COVID-19 cases and fatalities in each country, builds a behavior model of each time series, detects deviations from the expected behavior, analyses the data, generates insights, and predicts the future. 


Data visualization is key to understanding data, but when you have many data series, e.g., the number of COVID-19 cases and the number of fatalities per country, looking at each of them is a tedious, time-consuming task. Yokaranda takes on the burden of this, learns the behavior of each country, detects deviations from its expected behavior, analyses the data, and generates insights. 

The Results


In the graph below, the blue dots represent the actual values reported, the purple line the expected values, and the purple area the expected margin of the prediction. On the right part of the graph, you can see an area without blue dots; this is the prediction.

Yokaranda is built to scale. In this COVID-19 case, we have around 600-time series, a volume that can be analyzed by data scientists; however, if you had thousands or millions of time series, for example data of a massive amount of Smart IoT devices, the manual analysis would be impossible.

Comparing COVID-19 Behavior Across Countries

Governments are trying to learn from the experience of other counties in their efforts to fight COVID-19.


Yokaranda helps to automatically identify similar COVID-19 behavior across countries and learn from their experiences before taking action. In this example, Portugal has similar COVID-19 behavior as seen in the Dominican Republic. You can expect that if Portugal gains positive results from some restrictive actions or vaccination, similar results may be expected in the Dominican Republic.


Moreover, insights can also be gained from the actions taken in a specific country. For example, you can see that in Israel that the surge in COVID-19 cases occurred on September 1st and then reoccurred on September 8th, which should be worthwhile investigating. Yokaranda's machine learning capabilities automatically improve its  models and the results by learning from the user's feedback.


Yokaranda detected that the behavior of Israel in the last seven days is similar to the behavior of Germany (21 days ago) and of the Netherlands (30 days ago).

Screenshot 2021-10-20 232105.jpg





Yokaranda offers a powerful monitoring solution to massive data sets, IoT, and Smart AI/ML IoT deployments, which is simple to use with your applications. In this case, we illustrated some of the benefits of analyzing COVID-19 publicly available data sources and provided some initial perspectives on the expected benefits in other industries, such as telecoms, manufacturing, and smart consumer devices.


For more details, please contact us at

Get In Touch

Thanks for submitting!

bottom of page