A New Approach For Constructing Missing Feature Values

Mohammad Ashraf Ali Bani Ahmad, Girija Chetty, Dat Tran, Dharmendra Sharma

    Research output: Contribution to journalArticle

    Abstract

    This paper presents a new approach for constructing missing features values based on iterative nearest neighbors and distance metrics. The proposed approach employs weighted k nearest neighbors' algorithm and propagating the classification accuracy to a certain threshold. The proposed method showed improvement of classification accuracy of 0.005 in the constructed dataset than the original dataset which contain some missing feature values. The maximum classification accuracy was 0.9698 on k=1. This work is a component from a research for an automated diagnosing for breast cancer. The main aim of the current paper is to prepare the dataset for mining process. Future works includes applying the proposed method on more datasets and apply feature selection tools to reduce attributes number.
    Original languageEnglish
    Pages (from-to)110-118
    Number of pages9
    JournalInternational Journal of Intelligent Information Processing
    Volume3
    Issue number1
    DOIs
    Publication statusPublished - 2012

    Fingerprint

    Feature extraction

    Cite this

    @article{b8f9cf74ceaa44fbac2714966463ce22,
    title = "A New Approach For Constructing Missing Feature Values",
    abstract = "This paper presents a new approach for constructing missing features values based on iterative nearest neighbors and distance metrics. The proposed approach employs weighted k nearest neighbors' algorithm and propagating the classification accuracy to a certain threshold. The proposed method showed improvement of classification accuracy of 0.005 in the constructed dataset than the original dataset which contain some missing feature values. The maximum classification accuracy was 0.9698 on k=1. This work is a component from a research for an automated diagnosing for breast cancer. The main aim of the current paper is to prepare the dataset for mining process. Future works includes applying the proposed method on more datasets and apply feature selection tools to reduce attributes number.",
    author = "{Bani Ahmad}, {Mohammad Ashraf Ali} and Girija Chetty and Dat Tran and Dharmendra Sharma",
    year = "2012",
    doi = "10.4156/IJIIP.vol3.issue1.11",
    language = "English",
    volume = "3",
    pages = "110--118",
    journal = "International Journal of Intelligent Information Processing",
    issn = "2093-1964",
    number = "1",

    }

    A New Approach For Constructing Missing Feature Values. / Bani Ahmad, Mohammad Ashraf Ali; Chetty, Girija; Tran, Dat; Sharma, Dharmendra.

    In: International Journal of Intelligent Information Processing, Vol. 3, No. 1, 2012, p. 110-118.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - A New Approach For Constructing Missing Feature Values

    AU - Bani Ahmad, Mohammad Ashraf Ali

    AU - Chetty, Girija

    AU - Tran, Dat

    AU - Sharma, Dharmendra

    PY - 2012

    Y1 - 2012

    N2 - This paper presents a new approach for constructing missing features values based on iterative nearest neighbors and distance metrics. The proposed approach employs weighted k nearest neighbors' algorithm and propagating the classification accuracy to a certain threshold. The proposed method showed improvement of classification accuracy of 0.005 in the constructed dataset than the original dataset which contain some missing feature values. The maximum classification accuracy was 0.9698 on k=1. This work is a component from a research for an automated diagnosing for breast cancer. The main aim of the current paper is to prepare the dataset for mining process. Future works includes applying the proposed method on more datasets and apply feature selection tools to reduce attributes number.

    AB - This paper presents a new approach for constructing missing features values based on iterative nearest neighbors and distance metrics. The proposed approach employs weighted k nearest neighbors' algorithm and propagating the classification accuracy to a certain threshold. The proposed method showed improvement of classification accuracy of 0.005 in the constructed dataset than the original dataset which contain some missing feature values. The maximum classification accuracy was 0.9698 on k=1. This work is a component from a research for an automated diagnosing for breast cancer. The main aim of the current paper is to prepare the dataset for mining process. Future works includes applying the proposed method on more datasets and apply feature selection tools to reduce attributes number.

    U2 - 10.4156/IJIIP.vol3.issue1.11

    DO - 10.4156/IJIIP.vol3.issue1.11

    M3 - Article

    VL - 3

    SP - 110

    EP - 118

    JO - International Journal of Intelligent Information Processing

    JF - International Journal of Intelligent Information Processing

    SN - 2093-1964

    IS - 1

    ER -