The world of technology is a fascinating one and it never ceases to amaze and surprise you. The things that we never thought possible are increasingly becoming reality and forcing us to think if the impossible of today is really just a possibility waiting to happen.
A popular saying goes, “It is always considered impossible until it’s done.” This has been quite true of the many advancements in several fields of technology that we have recently witnessed.
Take an example of the field of computer science which is rapidly changing our views on ‘computers’ and what they can do. We are now entering an era where computers are going to take on an identity of their own and do our work for us like never before.
Things like machine learning, predictive modelling, neural networks, artificial intelligence have all opened up the doors to totally new sets of abilities that we can use in various ways but today I’m going to talk about how we can use computer algorithms today to predict something on a TV show!
Game of Thrones
There is a high chance you already watch the wildly popular TV show ‘Game of Thrones’ or that at least you are aware of it. If you haven’t watched it already, stop reading this and go do watch the show already! It’s super awesome.
It is basically a show about this fantasy world that resembles the middle ages and is based on the interplay of several houses and independent characters that want the control over the entire kingdom and sit on the Iron Throne. Hence the name of the show.
That’s not all though, as the show contains magic, fire breathing dragons, epic wars, ice zombies, gory deaths, betrayal, and a lot more. Basically this show combines all the different cool stuff from different TV shows and combines them into one epic story.
One of the central elements and probably the most loved thing about the show is that no character, no matter how beloved, is safe and may get killed at any time. We were all in shock when the fan favourite character Ned Stark was killed off in the first season alone, who many people believed, was going to be the ultimate hero.
With so many important and beloved characters and the sword of uncertainty hanging over their heads at all times, we all wish we had a way of knowing which of our favourite ones are safe and which aren’t, especially now that only two seasons are left. Well, thanks to some researchers, it might be possible now.
A Game of Predictions
Image Credit: Milan Janosov
Milan Janosov is a Ph.D. student at the Centre for Network Science at the Central European University and he wanted to predict the fate of the characters so much, that he built a computer algorithm to that for him.
He basically built a social network map (image above) of the fictional land of Westeros, based on how different characters interact with each other, their social positions, their true friends and other parameters and used machine learning methods to predict their probability of survival.
He has even come up with the following chart with the most important characters sorted by increasing probability of survival.
Image Credit: Milan Janosov
As you can see, the algorithm lists Tyene Sand, as having the highest probability of death and you know what? She was poisoned two episodes ago and is almost certain to die! As the show progresses, we will have more information about just how accurate this algorithm really is.
A Testament To The Progress
Feats like this are testaments to the fact that computer systems are evolving and are capable of so much more these days than just a few years back. I mean who could have thought that we would be using them to predict deaths in TV shows!
As technology improves, we will be able to do so much more. Predictive models such as this are increasingly being integrated into AI systems to be able to make more accurate real world predictions.
It is already being used in health care for diagnosing diseases, in trading to make market predictions, in auto insurance, in marketing and so much more. In the future, we might just have real life oracles in the form of a computer!