ML Trends

Newsletter #5 - Apr 21, 2019

Winter is coming...

Winter is coming...

And it's true. Especially if you're living in the Southern Hemisphere where winter technically begins on June 1. But you know what we're talking about. And it wouldn't be appropriate if this week's newsletter wasn't inspired by Game of Thrones. So here goes nothin!

Will your favorite character survive?

Game of Thrones has entertained us on our television sets since 2011 and, after a 2-year wait, is back on for the final season.

The show has been notorious for killing off characters unexpectedly and most fans already have their predictions in place regarding the series-old question: At the end, who’s going to survive and who’s going to die?

In order to answer this question, a team comprising of Computer Science students at the Technical University of Munich have used machine learning to predict the survival rate of every GOT character.

The team gathered their data primarily from GOT-based Wikis and used a couple of approaches, i.e. Bayesian Inference with MCMC methods and a Neural Network approach, to create prediction models.

Disappointingly, Needle-wielding Arya has a 57% chance of surviving Season 8, while Cersei’s chances of survival are pretty high at 95%. Bronn, even with his wicked wit, is the most likely character to die with only a 6% chance of survival.

For more information, visit this link.

Coming to a theater near you!

Sticking with the current script of School, AI and Entertainment, another interesting use case comes from the Viterbi School of Engineering at USC.

An NC-17 rating from the MPAA is the most dreaded outcome for film producers. After spending millions of dollars on a big-budget film, a film no longer accessible for viewers under the age of 17 might mean expensive re-editing, reshooting or a decrease in box-office returns.

A research team from the Signal Analysis & Interpretation Lab at USC’s Viterbi School of Engineering has created a machine-learning based system that can predict the MPAA classification of a movie script. The team used a neural network approach to process natural language and focus on predicting the use of violence in scenes.

This system can help screenwriters, producers and directors predict the MPAA rating and help them make any required changes at an early stage of production.

For more information, visit this link.

Alexa, can you invent a new sport for me?

Ever wondered if AI could be used to create a sport? Of course you have! Or maybe you haven’t. Doesn’t matter, because that’s exactly what the team at AKQA, a design firm based in San Francisco, have done. Using RNNs and a DCGAN-based approach, AKQA has created a new sport called Speedgate.

Speedgate is a game played between two teams with 6 players each. The objective is to pass a Rugby-shaped ball through a gate. A game consists of three 7-minute periods. Players can hold on to the ball for 3 seconds before they have to pass or attempt a shot. Points are scored when a ball is successfully kicked through a gate, 2 points given to a regular goal and 3 points given to a Ricochet goal.

Researchers used rules and images from over 400 sports and tested numerous games before zeroing in on the rules of Speedgate.

For more information, visit this link.

We hope you enjoyed this edition of our newsletter. If you've got a couple of minutes to spare, do take this survey and help us get better.

If you found this interesting, you can help us grow by asking your friends, colleagues and enemies to subscribe to our newsletter.
Enjoy the rest of your weekend.

The ML Trends team
Copyright © 2019 mlTrends, All rights reserved.