Hierarchical Temporal Memory

Hierarchical Temporal Memory

Ein hierarchischer Temporalspeicher (Hierarchical Temporal Memory; HTM) ist ein Modell des maschinellen Lernens, welches von Jeff Hawkins und George Dileep (Numenta, Inc.) entwickelt wurde. Dieses Modell bildet einige Eigenschaften des Neocortex in einem Bayes'schen Netzwerk ab.

Inhaltsverzeichnis

Aufbau und Funktion

HTMs sind als hierarchisch aufgebautes Netz von Knoten organisiert. Jeder Knoten implementiert, wie in einem Bayes'schem Netzwerk, eine Lern- und Speicherfunktion. Die Struktur ist dahingehend aufgebaut, um anhand von zeitveränderlichen Daten eine hierarchische Präsentation dieser Daten zu erstellen. Dies ist jedoch nur möglich, wenn die Daten sowohl im (Problem-)Raum, als auch in der Zeit hierarchisch repräsentierbar sind.

Ein HTM führt die folgenden Funktionen aus, wobei die letzten zwei – je nach Implementierung – optional sind:

  1. Erkennung und Repräsentation von Elementen und Zusammenhängen.
  2. Inferenz von neuen Elementen und Zusammenhängen anhand der bekannten Elemente und Zusammenhängen.
  3. Erstellung von Voraussagen. Treffen diese Voraussagen nicht zu, wird das interne Modell entsprechend angepasst.
    siehe auch: Summen-Produkt-Algorithmus
  4. Verwenden von Voraussagen um Aktionen auszuführen, sowie Beobachtung der Auswirkung. (Steuerung und Regelung)

Kritik

HTMs stellen für KI-Forscher nichts grundlegend neues dar, sondern sind eine Kombination bereits vorhandener Techniken, wobei Jeff Hawkins jedoch nicht ausreichend auf die Ursprünge seiner Ideen verweist. Zudem hat Jeff Hawkins das in der Wissenschaft für Publikationen übliche Peer Review und damit eine fundierte Prüfung durch Wissenschaftler umgangen. Hierbei gilt es jedoch zu beachten, dass Hawkins nicht aus dem akademischen, sondern aus dem industriellen Umfeld stammt.

Siehe auch

  • universelle Wahrscheinlichkeitsmodelle
    • Bayes'sches Netzwerk
    • Energie basiertes Lernmodell
    • hierarchisches verstecktes Markov-Modell
    • Boltzmann Maschine
    • Helmholtz Maschine
  • nichtgenerative Modelle
  • empirische neurobiologische Modelle
    • HMAX-Modell

Literatur

Weblinks


Wikimedia Foundation.

Игры ⚽ Нужно решить контрольную?

Schlagen Sie auch in anderen Wörterbüchern nach:

  • Hierarchical Temporal Memory — (HTM) is a machine learning model developed by Jeff Hawkins and Dileep George of Numenta, Inc. that models some of the structural and algorithmic properties of the neocortex using an approach somewhat similar to Bayesian networks. HTM model is… …   Wikipedia

  • Memory-prediction framework — The memory prediction framework is a theory of brain function that was created by Jeff Hawkins and described in his 2004 book On Intelligence. This theory concerns the role of the mammalian neocortex and its associations with the hippocampus and… …   Wikipedia

  • Hierarchical hidden Markov model — The Hierarchical hidden Markov model (HHMM) is a statistical model derived from the hidden Markov model (HMM). In an HHMM each state is considered to be a self contained probabilistic model. More precisely each stateof the HHMM is itself an HHMM …   Wikipedia

  • Memory hierarchy — Diagram of the computer memory hierarchy See also: Computer data storage The term memory hierarchy is used in the theory of computation when discussing performance issues in computer architectural design, algorithm predictions, and the lower… …   Wikipedia

  • Music-related memory — Musical memory refers to the ability to remember music related information, such as melodic content and other progressions of tones or pitches. The differences found between linguistic memory and musical memory have led researchers to theorize… …   Wikipedia

  • Semantic memory — refers to the memory of meanings, understandings, and other concept based knowledge unrelated to specific experiences. The conscious recollection of factual information and general knowledge about the world,cite web… …   Wikipedia

  • On Intelligence —   …   Wikipedia

  • Artificial intelligence — AI redirects here. For other uses, see Ai. For other uses, see Artificial intelligence (disambiguation). TOPIO, a humanoid robot, played table tennis at Tokyo International Robot Exhibition (IREX) 2009.[1] Artificial intelligence ( …   Wikipedia

  • Neural decoding — is a neuroscience related field concerned with the reconstruction of sensory and other stimuli from information that has already been encoded and represented in the brain by networks of neurons. The main goal of studying neural decoding is to… …   Wikipedia

  • List of computing and IT abbreviations — This is a list of computing and IT acronyms and abbreviations. Contents: 0–9 A B C D E F G H I J K L M N O P Q R S T U V W X Y …   Wikipedia

Share the article and excerpts

Direct link
Do a right-click on the link above
and select “Copy Link”