Video Experimental Relacionado
Updated: Aug 26, 2025

13:19
Deep Neural Networks for Image-Based Dietary Assessment
Published on: March 13, 2021
9.3K
Descubrir algoritmos de multiplicación de matrices más rápidos con aprendizaje por refuerzo
Alhussein Fawzi1, Matej Balog2, Aja Huang2
1DeepMind, London, UK. afawzi@deepmind.com.
Nature
|October 5, 2022
Resumen
El aprendizaje por refuerzo profundo, a través de AlphaTensor, descubre nuevos algoritmos de multiplicación de matrices. Este enfoque de IA mejora significativamente la eficiencia computacional, superando los métodos diseñados por humanos para tamaños de matriz clave.
Área de la Ciencia:
- Ciencias de la computación
- Inteligencia artificial
- Matemáticas computacionales
Sus antecedentes:
- La multiplicación matricial es un cálculo fundamental que afecta a diversos campos como las redes neuronales y la computación científica.
- Descubrir nuevos algoritmos para la multiplicación de matrices es un desafío debido al vasto espacio de búsqueda.
- Los algoritmos existentes, aunque eficientes, pueden no representar la solución óptima.
Objetivo del estudio:
- Desarrollar un enfoque impulsado por la IA para descubrir algoritmos de multiplicación de matrices eficientes y comprobadamente correctos.
- Explorar el potencial del aprendizaje por refuerzo profundo en la automatización del descubrimiento algorítmico.
- Para lograr avances en la complejidad de la multiplicación de matrices más allá de la intuición humana.
Principales métodos:
- Utilizó un agente de aprendizaje de refuerzo profundo, AlphaTensor, inspirado en AlphaZero.
- Entrenó a AlphaTensor para jugar un juego centrado en encontrar descomposiciones tensoriales dentro de un espacio de factores finito.
- Aplicó el agente para descubrir algoritmos para la multiplicación de matrices arbitraria y estructurada.
Principales resultados:
- AlphaTensor descubrió algoritmos que superan la complejidad del estado de la técnica para varias dimensiones de la matriz.
- Un nuevo algoritmo para matrices 4x4 en un campo finito mejora el método de Strassen de 50 años.
- Optimización demostrada para tiempos de ejecución de hardware específicos y multiplicación de matrices estructuradas.
Conclusiones:
- El aprendizaje de refuerzo profundo, ejemplificado por AlphaTensor, puede acelerar el descubrimiento algorítmico.
- El enfoque ofrece un camino para superar los algoritmos diseñados por humanos para tareas computacionales fundamentales.
- AlphaTensor proporciona un marco flexible para optimizar algoritmos basados en diferentes criterios, incluida la complejidad computacional y la eficiencia práctica.
Videos de Conceptos Relacionados
Observational Learning
281
Albert Bandura's observational learning, also known as imitation or modeling, occurs when a person observes and imitates another's behavior. It is a quicker process than operant conditioning. A well-known example is the Bobo doll study, where children who saw an adult acting aggressively towards the doll were more likely to act aggressively when left alone, compared to those who observed a nonaggressive adult. Many psychologists view observational learning as a form of latent learning...
281
Reinforcement Schedules
231
Positive reinforcement is a powerful method for teaching new behaviors to both animals and humans. B.F. Skinner demonstrated this with his experiments using rats in a Skinner box. When a rat pressed a lever, it received a food pellet. This immediate reward encouraged the rat to repeat the behavior. This method, where a reward follows every instance of the behavior, is known as continuous reinforcement. It is highly effective for establishing new behaviors quickly.
Once a behavior is learned,...
Once a behavior is learned,...
231
Reinforcement
313
Positive and negative reinforcement are key concepts in operant conditioning, a learning process where the consequences of a behavior affect the likelihood of that behavior being repeated.
Positive reinforcement occurs when a behavior is followed by the presentation of a rewarding stimulus, increasing the frequency of that behavior. For example:
Positive reinforcement occurs when a behavior is followed by the presentation of a rewarding stimulus, increasing the frequency of that behavior. For example:
313
Associative Learning
525
Associative learning is a fundamental concept in behavioral psychology, wherein a connection is established between two stimuli or events, leading to a learned response. This process is critical in understanding how behaviors are acquired and modified. Conditioning, the mechanism through which associations are formed, can be divided into two main types: classical conditioning and operant conditioning, each elucidating different aspects of associative learning.
Classical conditioning, also known...
Classical conditioning, also known...
525
Machines: Problem Solving II
355
Machines are complex structures consisting of movable, pin-connected multi-force members that work together to transmit forces. Consider a lifting tong carrying a 100 kg load. It comprises movable sections DAF and CBG linked together with member AB.
355
Ampere-Maxwell's Law: Problem-Solving
730
A parallel-plate capacitor with capacitance C, whose plates have area A and separation distance d, is connected to a resistor R and a battery of voltage V. The current starts to flow at t = 0. What is the displacement current between the capacitor plates at time t? From the properties of the capacitor, what is the corresponding real current?
To solve the problem, we can use the equations from the analysis of an RC circuit and Maxwell's version of Ampère's law.
For the first part of...
To solve the problem, we can use the equations from the analysis of an RC circuit and Maxwell's version of Ampère's law.
For the first part of...
730

