MACHINE LEARNING FOR VOTER CLUSTERING
Since the beginning of the century, technological capacity of mankind has increased significantly. Until recently, only the military and IT giants such as Google, Apple and Microsoft have possessed such resources. Over the past decade, these companies have amounted such impressive capacities that they have opened access to them to the general public. Today, you can rent a computing cluster from Google, Amazon or Microsoft for quite a reasonable price. Moreover, a few years ago, Google openly posted the algorithms it uses to solve its management problems. This has drastically decreased the financial threshold required to exploit the advantages this technology provides. Consequently, there is widespread use of big data technologies in various spheres of society, including politics.
Clustering the electoral field based on the lists of voters, their location, and the data from past elections allows the political parties to radically increase the productivity of their work. Moreover, surveys conducted by volunteers via the mobile application can be an important source of data for machine learning algorithms.
Technologies for Big Data analysis help to see the whole picture and make conclusions in the most accessible way for everyone. An additional advantage of implementing such technologies is the ability to effectively observe possible falsifications in the polling stations and predict the results of elections.