**Series Number**S029

**Source**PIRSA

The adoption of machine learning (ML) into theoretical physics comes on the heels of an explosion of industry progress that started in 2012. Since that time, computer scientists have demonstrated that learning algorithms - those designed to respond and adapt to new data - provide an exceptionally powerful platform for tackling many difficult tasks in image recognition, natural language comprehension, game play and more. This new breed of ML algorithm has now conquered benchmarks previously thought to be decades away due to their high mathematical complexity. In the last several years, researchers at Perimeter have begun to examine machine learning algorithms for application to a new set of problems, including condensed matter, quantum information, numerical relativity, quantum gravity and astrophysics.

**Series Number**S029

**Source**PIRSA

The adoption of machine learning (ML) into theoretical physics comes on the heels of an explosion of industry progress that started in 2012. Since that time, computer scientists have demonstrated that learning algorithms - those designed to respond and adapt to new data - provide an exceptionally powerful platform for tackling many difficult tasks in image recognition, natural language comprehension, game play and more. This new breed of ML algorithm has now conquered benchmarks previously thought to be decades away due to their high mathematical complexity. In the last several years, researchers at Perimeter have begun to examine machine learning algorithms for application to a new set of problems, including condensed matter, quantum information, numerical relativity, quantum gravity and astrophysics.