Mubarak # World Cup of Germany ... # Artificial intelligence reveals the holder of the title "# Russia 2018"
The MIT Technology Review website said that followers and interested people are trying very hard and unceasingly to predict the results of matches at any FIFA World Cup.
"There are many ways to measure the likely outcome of matches," the magazine's website said in a scientific article on the forecast of the results of the 2018 World Cup.
The article pointed out that betting companies use professional statisticians to analyze databases containing the probability of winning results for each team.
In this way, the betting company can offer the prospect of all matches in the coming weeks.
According to the companies, Brazil is the biggest candidate for the 2018 World Cup with 16.6 percent, followed by Germany with 12.8 percent and Spain with 12.5 percent.
In recent years, researchers have developed "machine learning" techniques that can outweigh traditional statistical methods.
In the expectations of these techniques, the answer comes from Andreas Grohl and his colleagues from the University of Dortmund, Germany, using a combination of "automated learning" and traditional statistics to create a model called the "random jungle" model to determine the most likely teams to win, World 100 thousand times.
They used 3 different ways to create models based on team performance in all previous games from 2002 to 2014.
Groll and his teammates used this method of modeling the 2018 World Cup as they designed the results of each game likely to be played by the teams and used the results to build the most likely track in the tournament.
According to the study, Germany's chances of reaching the quarter-finals are up to 58 percent. By contrast, Spain are unlikely to face a strong 16-round clash, which means they have a 73 percent chance of reaching the quarter-finals.
"Spain have the best chances to win," said Groll. "But if Germany can reach the quarterfinals, it will be in first place