Selected article for: "feature vector and sae output"

Author: Saban Ozturk; Umut Ozkaya; Mucahid Barstugan
Title: Classification of Coronavirus Images using Shrunken Features
  • Document date: 2020_4_6
  • ID: 2l1zw19o_31
    Snippet: An AE architecture consists of two layers, encoder, and decoder, whose main purpose is to re-interpret the relationship between input and output. In the proposed study, AE architecture expresses the high-dimensional feature vector with fewer parameters. AE architecture, trained in an unsupervised style, creates a relationship between input and output. In AE architecture, more than one AE is added in series. Let X={xn} N be an N feature vector. Ne.....
    Document: An AE architecture consists of two layers, encoder, and decoder, whose main purpose is to re-interpret the relationship between input and output. In the proposed study, AE architecture expresses the high-dimensional feature vector with fewer parameters. AE architecture, trained in an unsupervised style, creates a relationship between input and output. In AE architecture, more than one AE is added in series. Let X={xn} N be an N feature vector. New shrunken feature vectors to be obtained at the sAE output F:X→[0,1] Nxk , is a k-bit vector (bnϵ[0,1] k ) for each feature vector xn.

    Search related documents:
    Co phrase search for related documents
    • feature vector and propose study: 1
    • output input and propose study: 1, 2