November 4, 2019 Bachelor Thesis Open Access

Heine, Greta Sophie

JSON-LD (schema.org) Export

{
  "@context": "https://schema.org/", 
  "@type": "ScholarlyArticle", 
  "contributor": [], 
  "creator": [
    {
      "@type": "Person", 
      "affiliation": "KIT/ETP", 
      "name": "Heine, Greta Sophie"
    }
  ], 
  "datePublished": "2019-11-04", 
  "description": "<p>This thesis enhances the understanding of Neural Network (NN) trainings by<br>\ninvestigation of the learning process especially focusing on the dependence<br>\nof the NN output on the input space for given tasks. For this purpose, the<br>\nNN function is decomposed into a Taylor expansion. The Taylor coefficients<br>\nserve as a metric to illustrate the influence of input space features on the<br>\noutput at each step of the training. Both, the arithmetic mean values of the<br>\nTaylor coefficients and their dependence on each point of the input space are<br>\ninvestigated, giving new insights into the decision taking of NNs.</p>", 
  "headline": "Illustration of the Neural Network Learning Process during Training", 
  "image": "https://publish.etp.kit.edu/static/img/logos/zenodo-gradient-round.svg", 
  "inLanguage": {
    "@type": "Language", 
    "alternateName": "eng", 
    "name": "English"
  }, 
  "keywords": [
    "neural network, training, Taylor expansion"
  ], 
  "name": "Illustration of the Neural Network Learning Process during Training", 
  "url": "https://publish.etp.kit.edu/record/21984"
}