How statistical invariances will help us build AI systems exhibiting human-like performance by following human-like strategies.
Current machine learning systems crumble when the distributions of training and testing examples differ in spurious correlations. This is a major roadblock toward the development of advanced machine intelligence, which demands not only human-like performance but the deployment of human-like strategies. The prevalent approach in AI, fixated on recklessly minimizing average training error, falls short in producing AI systems capable of authentic out-of-distribution generalization. This book introduces the Invariance Principle, a new epistemological tool to unearth correlations invariant across diverse collections of empirical data.
The Invariance Principle, encapsulated in the axiom “frame your problem so its answer matches across circumstances," will not only find its practical incarnation in the family of Invariant Risk Minimization algorithms, but also illuminate our understanding of causation. It will permeate topics such as environment discovery, large-language models, self-supervised learning, mixing data augmentation, uncertainty estimation, and fairness. The author argues that the Invariance Principle is a central inductive bias fueling advances across fields of knowledge, such as physics, metaphysics, and cognitive science.
The final chapter includes personal examples of how invariance has shaped the author’s understanding of his own subjective experience, as well as how he has interpreted both Eastern and Western philosophical traditions.
Les informations fournies dans la section « Synopsis » peuvent faire référence à une autre édition de ce titre.
David Lopez-Paz is a research scientist at FAIR, Meta. Previously, he held positions in the European Space Agency, RedBull, Formula 1, and Google Research.
Les informations fournies dans la section « A propos du livre » peuvent faire référence à une autre édition de ce titre.
Vendeur : Books From California, Simi Valley, CA, Etats-Unis
paperback. Etat : Fine. N° de réf. du vendeur mon0004162205
Quantité disponible : 1 disponible(s)
Vendeur : PBShop.store UK, Fairford, GLOS, Royaume-Uni
PAP. Etat : New. New Book. Shipped from UK. Established seller since 2000. N° de réf. du vendeur GB-9780262053341
Quantité disponible : 5 disponible(s)
Vendeur : Grand Eagle Retail, Bensenville, IL, Etats-Unis
Paperback. Etat : new. Paperback. How statistical invariances will help us build AI systems exhibiting human-like performance by following human-like strategies.Current machine learning systems crumble when the distributions of training and testing examples differ in spurious correlations. This is a major roadblock toward the development of advanced machine intelligence, which demands not only human-like performance but the deployment of human-like strategies. The prevalent approach in AI, fixated on recklessly minimizing average training error, falls short in producing AI systems capable of authentic out-of-distribution generalization. This book introduces the Invariance Principle, a new epistemological tool to unearth correlations invariant across diverse collections of empirical data.The Invariance Principle, encapsulated in the axiom "frame your problem so its answer matches across circumstances," will not only find its practical incarnation in the family of Invariant Risk Minimization algorithms, but also illuminate our understanding of causation. It will permeate topics such as environment discovery, large-language models, self-supervised learning, mixing data augmentation, uncertainty estimation, and fairness. The author argues that the Invariance Principle is a central inductive bias fueling advances across fields of knowledge, such as physics, metaphysics, and cognitive science.The final chapter includes personal examples of how invariance has shaped the author's understanding of his own subjective experience, as well as how he has interpreted both Eastern and Western philosophical traditions. Shipping may be from multiple locations in the US or from the UK, depending on stock availability. N° de réf. du vendeur 9780262053341
Quantité disponible : 1 disponible(s)
Vendeur : Rarewaves USA, OSWEGO, IL, Etats-Unis
Paperback. Etat : New. N° de réf. du vendeur LU-9780262053341
Quantité disponible : Plus de 20 disponibles
Vendeur : Rarewaves.com USA, London, LONDO, Royaume-Uni
Paperback. Etat : New. N° de réf. du vendeur LU-9780262053341
Quantité disponible : 5 disponible(s)
Vendeur : Revaluation Books, Exeter, Royaume-Uni
Paperback. Etat : Brand New. 9.00x6.00 inches. In Stock. N° de réf. du vendeur __0262053349
Quantité disponible : 1 disponible(s)
Vendeur : Majestic Books, Hounslow, Royaume-Uni
Etat : New. N° de réf. du vendeur 409556451
Quantité disponible : 3 disponible(s)
Vendeur : Brook Bookstore On Demand, Napoli, NA, Italie
Etat : new. N° de réf. du vendeur KYYVO48XPQ
Quantité disponible : 5 disponible(s)
Vendeur : Books Puddle, New York, NY, Etats-Unis
Etat : New. N° de réf. du vendeur 26404679228
Quantité disponible : 3 disponible(s)
Vendeur : Chiron Media, Wallingford, Royaume-Uni
paperback. Etat : New. N° de réf. du vendeur 6666-GRD-9780262053341
Quantité disponible : 3 disponible(s)