Healthcare & Data Science

Medical statistics is one of the key challenges facing the healthcare system. A table containing compressed and systematized data can provide feedback from the population, allowing doctors to analyze and understand which diseases are most prevalent and which pose a danger to the population. Choosing which data analysis method to use is often critical, but technology has greatly simplified this task.

‍

Innovative approach to influenza research

The Smorodintsev Research Institute of Influenza conducted a study on the spread of the influenza virus across Russia. Project managers had to decide which method to adopt and approached us with a request to process the compiled database. We collaborated with the institute to implement a healthcare data science project.

‍

Research Methods

We chose clustering as our data analysis method, which is applicable in studies from various fields. Clustering is particularly convenient in projects of this kind because it allows identification of independent clustering criteria. In conventional classification, objects are distributed into groups according to predetermined attributes. Among the attributes identified in our study were gender, age, and frequency of visits to clinics.

For the analysis algorithm, we chose Python because it is a very convenient tool in this case. Python already has many frameworks and libraries that optimize machine learning implementation. We selected K-means clustering, the most popular and fault-tolerant method, for data analysis.

‍

Results

We are proud to contribute to such a massive project and to make a significant impact on the Russian healthcare system by providing the Research Institute of Influenza with essential data regarding the prevalence of the influenza virus in Russia.