Gaussian-PSO with fuzzy reasoning based on structural learning for training a Neural Network

Haydee Melo, Junzo Watada

    Research output: Contribution to journalArticle

    30 Citations (Scopus)

    Abstract

    This paper proposes Gaussian-PSO-based structural learning and fuzzy reasoning to optimize the weights and the structure of the Feed Forward Neural Network. The Neural Network is widely used for various applications; though it still has disadvantages such as learning capability and slow convergence. Back Propagation, the most used learning algorithm, has several difficulties such as the necessity for a priori specification of the network structure and sensibility to parameter settings. Recently, research studies have introduced evolutionary algorithms into the learning to improve its performance. The PSO is a population-based algorithm that has the advantage of faster convergence. However, the total number of the weights in the Neural Network determines the size of each particle, therefore the size of the network structure is computationally time consuming. The proposed method improves the learning and removes the stress by eliminating the necessity of determining a detailed network.

    Original languageEnglish
    Pages (from-to)405-412
    Number of pages8
    JournalNeurocomputing
    Volume172
    DOIs
    Publication statusPublished - 2016 Jan 8

    Fingerprint

    Particle swarm optimization (PSO)
    Learning
    Neural networks
    Feedforward neural networks
    Backpropagation
    Evolutionary algorithms
    Learning algorithms
    Specifications
    Weights and Measures
    Particle Size
    Research
    Population

    Keywords

    • Fuzzy reasoning
    • GPSO
    • Neural networks
    • PSO
    • Structural learning

    ASJC Scopus subject areas

    • Artificial Intelligence
    • Computer Science Applications
    • Cognitive Neuroscience

    Cite this

    Gaussian-PSO with fuzzy reasoning based on structural learning for training a Neural Network. / Melo, Haydee; Watada, Junzo.

    In: Neurocomputing, Vol. 172, 08.01.2016, p. 405-412.

    Research output: Contribution to journalArticle

    @article{12e44fc9925e4d6dbe9fdaf1b4526a42,
    title = "Gaussian-PSO with fuzzy reasoning based on structural learning for training a Neural Network",
    abstract = "This paper proposes Gaussian-PSO-based structural learning and fuzzy reasoning to optimize the weights and the structure of the Feed Forward Neural Network. The Neural Network is widely used for various applications; though it still has disadvantages such as learning capability and slow convergence. Back Propagation, the most used learning algorithm, has several difficulties such as the necessity for a priori specification of the network structure and sensibility to parameter settings. Recently, research studies have introduced evolutionary algorithms into the learning to improve its performance. The PSO is a population-based algorithm that has the advantage of faster convergence. However, the total number of the weights in the Neural Network determines the size of each particle, therefore the size of the network structure is computationally time consuming. The proposed method improves the learning and removes the stress by eliminating the necessity of determining a detailed network.",
    keywords = "Fuzzy reasoning, GPSO, Neural networks, PSO, Structural learning",
    author = "Haydee Melo and Junzo Watada",
    year = "2016",
    month = "1",
    day = "8",
    doi = "10.1016/j.neucom.2015.03.104",
    language = "English",
    volume = "172",
    pages = "405--412",
    journal = "Neurocomputing",
    issn = "0925-2312",
    publisher = "Elsevier",

    }

    TY - JOUR

    T1 - Gaussian-PSO with fuzzy reasoning based on structural learning for training a Neural Network

    AU - Melo, Haydee

    AU - Watada, Junzo

    PY - 2016/1/8

    Y1 - 2016/1/8

    N2 - This paper proposes Gaussian-PSO-based structural learning and fuzzy reasoning to optimize the weights and the structure of the Feed Forward Neural Network. The Neural Network is widely used for various applications; though it still has disadvantages such as learning capability and slow convergence. Back Propagation, the most used learning algorithm, has several difficulties such as the necessity for a priori specification of the network structure and sensibility to parameter settings. Recently, research studies have introduced evolutionary algorithms into the learning to improve its performance. The PSO is a population-based algorithm that has the advantage of faster convergence. However, the total number of the weights in the Neural Network determines the size of each particle, therefore the size of the network structure is computationally time consuming. The proposed method improves the learning and removes the stress by eliminating the necessity of determining a detailed network.

    AB - This paper proposes Gaussian-PSO-based structural learning and fuzzy reasoning to optimize the weights and the structure of the Feed Forward Neural Network. The Neural Network is widely used for various applications; though it still has disadvantages such as learning capability and slow convergence. Back Propagation, the most used learning algorithm, has several difficulties such as the necessity for a priori specification of the network structure and sensibility to parameter settings. Recently, research studies have introduced evolutionary algorithms into the learning to improve its performance. The PSO is a population-based algorithm that has the advantage of faster convergence. However, the total number of the weights in the Neural Network determines the size of each particle, therefore the size of the network structure is computationally time consuming. The proposed method improves the learning and removes the stress by eliminating the necessity of determining a detailed network.

    KW - Fuzzy reasoning

    KW - GPSO

    KW - Neural networks

    KW - PSO

    KW - Structural learning

    UR - http://www.scopus.com/inward/record.url?scp=84947195900&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84947195900&partnerID=8YFLogxK

    U2 - 10.1016/j.neucom.2015.03.104

    DO - 10.1016/j.neucom.2015.03.104

    M3 - Article

    VL - 172

    SP - 405

    EP - 412

    JO - Neurocomputing

    JF - Neurocomputing

    SN - 0925-2312

    ER -