Nobel Prize in Physics awarded for training artificial neural networks (ANNs) using physics

Posted 09 Oct 2024

2 min read

It has been awarded jointly to John Hopfield & Geoffrey Hinton for constructing methods that helped lay foundation for machine learning (A type of AI) using ANNs.

What are ANNs?

ANNs are a subset of Machine Learning algorithms designed to model workings of human brain.
ANNs consist of interconnected nodes, or artificial neurons, that process information similarly to how neurons function in the human brain.

Discoveries

John Hopfield: He invented a type of neural network (Hopfield Network) which is designed to store and recall patterns, similar to how memory works.
- Hopfield network utilizes physics that describes a material’s characteristics due to its atomic spin.
  - Atomic spin is magnetic moment of an atom that is caused by spins of particles that make up atoms.
Geoffrey Hinton: He invented a method (Boltzman Machine) that can autonomously find properties in data e.g. identifying specific elements in pictures.
- Boltzmann machine learns by using examples that it may see while it works. It can sort images or create new patterns similar to what it learned.
  - This network uses methods from statistical physics.

Role of ANNs in AI

Deep Learning: ANNs are foundation of Deep Learning, a branch of Machine Learning that deals with large datasets and complex models.
Use in AI: AI systems used for image recognition, natural language processing, and autonomous systems rely on ANNs to make decisions without human intervention.
Learning from Data: For example, ANN trained on thousands of medical images can eventually detect tumors in new images with high accuracy.