Nobel Prize in Physics 2024 awarded for research in neural networks

The Nobel Committee recognized the contribution of two scientists to the development of artificial intelligence, having awarded They were awarded the 2024 Prize in Physics. John Hopfield and Geoffrey Hinton developed the methods that formed the basis of modern machine learning.

The laureates used physics tools to create powerful data processing algorithms. Hopfield invented associative memory capable of storing and recalling images and other types of information patterns. Hinton developed a method that autonomously detects properties in data, allowing for tasks such as identifying specific elements in images.

Artificial neural networks, inspired by the structure of the brain, consist of nodes that represent neurons. These nodes interact through connections like synapsesA synapse is a connection between neurons (or nodes in artificial neural networks) through which signals are transmitted., which can be strengthened or weakened. The network is trained, for example, by strengthening connections between nodes with simultaneously high values.

The Hopfield Network uses the method of pattern preservation and reconstruction. It is based on physical principles that describe material properties through atomic spinAtomic spin is a quantum mechanical term that describes the intrinsic angular momentum of an elementary particle such as an electron or an atom. . The network is trained by finding values for the connections between nodes so that the stored images are low energy. When the network is presented with a distorted or incomplete image, it updates the node values in stages, reducing the energy of the system to find the most similar stored image.

Jeffrey Hinton used the Hopfield network as the basis for a new network that uses a different method: Boltzmann machine. It can learn to recognize characteristic elements in this type of data. Hinton used tools from statistical physics — the science of systems built from many similar components. The machine learns by being presented with examples that are likely to arise in the course of the machine’s operation. A Boltzmann machine can be used to classify images or create new examples of the type of pattern it has been trained on. Hinton built on this work, helping to launch the current explosive development of machine learning.

The Nobel Prize in Physics has a rich history. It was founded by Alfred Nobel, a Swedish chemist and inventor of dynamite, in 1895. Past winners include outstanding scientists who have made revolutionary discoveries in quantum mechanics, nuclear physics, astrophysics, and other fields.