Talk

How to train machine learning agents to do fun and absurd things using Unity3D.
Conference (INTERMEDIATE level)
Room 9

While machine learning is often used for image recognition, speech recognition and the such, people tend to forget that there is another industry making full use of machine learning. The games industry. Within the games industry companies have been using machine learning to train their NPCs (non-player characters) a plethora of things. From basic combat behavior to full blown city building. During this session Alexander will show you what you can expect from using machine learning agents in Unity3D, an industry standard game engine. Alexander will demonstrate how to implement basic machine learning agents in Unity and how to train them! But the agents won't be trained to do standard stuff. From matrix bullet dodging to headbanging, we can train our agents to do the most absurd and fun things!

Alexander Chatzizacharias
JDriven

Alexander is a 31-year old Software Engineer at jDriven. He is of Dutch/Greek nationality. Alexander has a bachelor in Game Development (University of applied sciences of Amsterdam) and a master in Game studies (University of Amsterdam). During his master Alexander developed a passion for gamification and software engineering. One of Alexanders career goals is bringing game development and software engineering closer together. He believes that both industries can learn a lot from each other, and is continuously trying to mix technologies and methodologies from both fields.