Video Experimental Relacionado
Updated: Feb 20, 2026
13:40
Combining Computer Game-Based Behavioural Experiments With High-Density EEG and Infrared Gaze Tracking
Published on: December 16, 2010
17.3K
Dominar el juego de Go sin conocimiento humano
David Silver1, Julian Schrittwieser1, Karen Simonyan1
1DeepMind, 5 New Street Square, London EC4A 3TW, UK.
Nature
|October 21, 2017
Resumen
Un nuevo algoritmo de inteligencia artificial, AlphaGo Zero, aprende únicamente a través del aprendizaje por refuerzo sin datos humanos. Logró un rendimiento sobrehumano en el juego de Go entrenando con sus propios datos de juego.
Área de la Ciencia:
- Inteligencia artificial
- Aprendizaje por refuerzo
- Teoría de juegos
Sus antecedentes:
- Un objetivo clave en la inteligencia artificial (IA) es crear algoritmos capaces de lograr una competencia sobrehumana en dominios complejos.
- Los sistemas de IA anteriores, como AlphaGo, utilizaban el aprendizaje supervisado de datos de expertos humanos y el aprendizaje de refuerzo del autojuego.
- Sin embargo, estos sistemas todavía dependían del conocimiento y la guía humanos.
Objetivo del estudio:
- Desarrollar un algoritmo de IA que aprenda exclusivamente a través del aprendizaje por refuerzo, sin ningún dato humano o experiencia en el dominio.
- Demostrar que una IA puede convertirse en su propio maestro, mejorando su rendimiento de forma iterativa a través del autojuego.
- Para lograr un rendimiento sobrehumano en el juego de Go utilizando un nuevo enfoque de IA autodidacta.
Principales métodos:
- El estudio presenta un algoritmo, AlphaGo Zero, basado únicamente en el aprendizaje por refuerzo.
- Una red neuronal está entrenada para predecir las propias selecciones de movimiento de AlphaGo Zero y los resultados del juego.
- Esta red neuronal mejora el algoritmo de búsqueda de árboles, lo que lleva a una mejor selección de movimientos y un autojuego más fuerte en iteraciones posteriores.
Principales resultados:
- AlphaGo Zero aprendió completamente desde cero, sin datos humanos, orientación o conocimiento de dominio más allá de las reglas del juego.
- El algoritmo logró un rendimiento sobrehumano en el juego de Go.
- AlphaGo Zero derrotó al anterior campeón por una puntuación de 100-0.
Conclusiones:
- El aprendizaje por refuerzo solo, sin datos humanos, es suficiente para que una IA logre un rendimiento sobrehumano en dominios desafiantes.
- El autojuego y la auto-mejora iterativa son mecanismos poderosos para el aprendizaje y el desarrollo de la IA.
- AlphaGo Zero representa un avance significativo en la IA, demostrando un nuevo paradigma para el entrenamiento de algoritmos.
Videos de Conceptos Relacionados
01:12Observational Learning
1.1K
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...
1.1K
01:14Understanding Deception
183
Deception is a pervasive aspect of human communication. Empirical studies have shown that most individuals engage in some form of deceit on a daily basis, with approximately 20% of social exchanges involving deceptive elements. Lying follows a developmental trajectory, peaking during adolescence and declining with age, possibly due to the maturation of cognitive control and social accountability.Cognitive and Social Factors in Deception DetectionDespite its prevalence, accurately detecting...
183
01:22Purposive Learning
535
E. C. Tolman emphasized the purposiveness of behavior — the idea that much of our behavior is goal-directed. For instance, employees who aim for a promotion work diligently to meet their targets. Tolman argued that when classical conditioning and operant conditioning occur, the organism acquires certain expectations. In classical conditioning, a child might fear a dog because they expect it to bite. In operant conditioning, a person might consistently work overtime because they expect a...
535
01:19High-Level and Low-Level Awareness
799
Controlled processes in human consciousness represent high-alert mental states where individuals deliberately focus their attention on achieving specific goals. Controlled processes can be seen in situations like mastering new technology, where a person might become so absorbed that they ignore surrounding distractions. Such processes involve selective attention, requiring one to concentrate on particular elements of experience while disregarding others. These are governed by executive...
799
01:14Avoidance Learning and Learned Helplessness
2.7K
Avoidance learning and learned helplessness are critical concepts in understanding behavioral responses to negative stimuli.
Avoidance learning occurs when an organism learns that a specific behavior can prevent an unpleasant outcome. For example, a student who receives a bad grade may start studying harder to avoid future poor grades. This behavior persists even when the negative outcome is no longer present. Avoidance learning is powerful because it maintains behavior in the absence of the...
Avoidance learning occurs when an organism learns that a specific behavior can prevent an unpleasant outcome. For example, a student who receives a bad grade may start studying harder to avoid future poor grades. This behavior persists even when the negative outcome is no longer present. Avoidance learning is powerful because it maintains behavior in the absence of the...
2.7K
01:30Machines: Problem Solving II
689
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.
689