Modeling cortical functionality with Hierarchical Temporal Memory

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F15%3A00234052" target="_blank" >RIV/68407700:21230/15:00234052 - isvavai.cz</a>

Výsledek na webu

<a href="http://www.researchgate.net/publication/283056327_Modeling_cortical_functionality_with_Hierarchical_Temporal_Memory_%28applications__extending_human_cognitive_capabilities%29" target="_blank" >http://www.researchgate.net/publication/283056327_Modeling_cortical_functionality_with_Hierarchical_Temporal_Memory_%28applications__extending_human_cognitive_capabilities%29</a>

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Modeling cortical functionality with Hierarchical Temporal Memory

Popis výsledku v původním jazyce

Inspired by neocortex, HTM should be suitable for similar high-level cognitive tasks, mainly for streaming tem- poral sequence data processing (sensory data) and anomaly detection. The HTM and its learning process have some implications so we would liketo define a class of tasks that are more suited for a HTM then other tasks or other ANN architectures. Further, we would like to apply a neural network models to some biological experiments, and therefor biological structures/processes need to be modeledwith an appropriate details, and still be usable for the machine learning tasks. Anomaly detection is a task with a very wide range of applications in the modern world (security, smart cities, medical), and is in principle inherently present in HTM. Wefound a suitable, well known and annotated dataset on ECG Arrhythmia ([3]) and perform one-class anomaly detection task on it. In the end, I try to simplify the current process of carrying out experiments on the NuPIC platform [6] and eva

Název v anglickém jazyce

Modeling cortical functionality with Hierarchical Temporal Memory

Popis výsledku anglicky

Inspired by neocortex, HTM should be suitable for similar high-level cognitive tasks, mainly for streaming tem- poral sequence data processing (sensory data) and anomaly detection. The HTM and its learning process have some implications so we would liketo define a class of tasks that are more suited for a HTM then other tasks or other ANN architectures. Further, we would like to apply a neural network models to some biological experiments, and therefor biological structures/processes need to be modeledwith an appropriate details, and still be usable for the machine learning tasks. Anomaly detection is a task with a very wide range of applications in the modern world (security, smart cities, medical), and is in principle inherently present in HTM. Wefound a suitable, well known and annotated dataset on ECG Arrhythmia ([3]) and perform one-class anomaly detection task on it. In the end, I try to simplify the current process of carrying out experiments on the NuPIC platform [6] and eva

Druh

O - Ostatní výsledky

CEP obor

IN - Informatika

OECD FORD obor

—

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2015

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů