The series “Game of thrones” is a real phenomenon. Six full seasons (recently saw the premiere of the 7th season). Millions of fans around the world. Dal the tears shed for the beloved, but so unfair to the killed characters. It can rightly be called one of the most iconic series of recent times. It is therefore hardly surprising that this socio-cultural phenomenon and interested people in the world of science.
Graduate of the Center for network science Central European University and Milan Anosov created an algorithm with 400 nodes and 3000 edges, which determines the probability of death the characters of the last two seasons of the popular series in the world of “Game of thrones” produced by HBO and is known for its “style” to kill liked by the audience characters.
To get the working model of the probability of Anosov used “seven different network functions that determine various aspects of social significance” for each individual character of the series, which he received from the analysis of scenes in which a graduate student gives the definition of “basic units of social interaction”. He further combined this analysis with the studied characteristic features of the previously killed characters, and on this basis derived the probability according to which a given character will survive till the end of this television show, is scheduled somewhere in 2018-2019.
For example, the model likelihood was determined that the Tien sand, daughter of Oberyn Martell, who suffered a cruel end of season 4 of the series, most likely, too will die, while Theon Greyjoy, who have experienced horrific torture from the hands of Ramsay Bolton will die with a probability of only 0.05 percent.
Perhaps this work can be called one of the most important, popular and vivid examples of what level reached the field of predictive modeling.
“A team of researchers who built the model usually works by finding answers to complex biological questions through datamining and machine learning algorithms”, says Dr. guy Yachdav involved in project management.
“The TV series “Game of thrones” is a great platform for research because it provides a virtually endless source of the raw multidimensional data,” said Ahdab on the website Phys.org.
The full list of predictions can be found on the website of the Center for network Sciences.
Created an AI algorithm that predicts the death of a character “Game of thrones”
Nikolai Khizhnyak