Tame Any Sequence with Transformers

Sequences are important in machine learning because they allow models to learn the relationships between items and predict the next ones.

Tarek Samaali
9 min readMay 3, 2022
Photo by Bradyn Trollip from Unsplash

This is important for time-series prediction and natural language processing tasks.

In this article, I will give a simple recipe that goes beyond what Transformers are known to achieve and predict a specific example of sequences that I take from the gaming world.

Tekken 7 and Artificial Intelligence:

Some possible applications of machine learning in gaming could include automatic game balancing (based on player data), better AI opponents, and improved game recommendations.

Tekken 7 is a fighting video game developed and published by Bandai Namco Entertainment. The game is the ninth installment in the Tekken series and the first to make use of the Unreal Engine.

In terms of artificial intelligence, Tekken 7 features several improvements over its predecessors. These include the ability to identify and react to the player’s tendencies, as well as increased variation in attack and defense.

I find Tekken 7 interesting because the developers took the time to explore the characters’…

--

--

Tarek Samaali

Senior Data Scientist @ Intermarché ( France ). Bookaholic. Metalhead. Former musician. Passionate about Data, Software & Technology.