Select Language





528 浏览
0 喜欢
2 次下载
0 条讨论
Life Classification

Standardized version of the original audiology database......

数据结构 ? 0M

    Data Structure ?

    * 以上分析是由系统提取分析形成的结果,具体实际数据为准。

    Original Version:

    (a) Original Owner: Professor Jergen at Baylor College of Medicine
    (b) Donor: Bruce Porter (porter '@' fall.cs.utexas.EDU)

    Standardized Version:

    (a) Donor: Ross Quinlan

    Data Set Information:

    This database is a standardized version of the original audiology database (see audiology.* in this directory). The non-standard set of attributes have been converted to a standard set of attributes according to the rules that follow.

    * Each property that appears anywhere in the original .data or .test file has been represented as a separate attribute in this file.

    * A property such as age_gt_60 is represented as a boolean attribute with values f and t.

    * In most cases, a property of the form x(y) is represented as a discrete attribute x() whose possible values are the various y's; air() is an example. There are two exceptions:
    ** when only one value of y appears anywhere, e.g. static(normal). In this case, x_y appears as a boolean attribute.
    ** when one case can have two or more values of x, e.g. history(..). All possible values of history are treated as separate boolean attributes.

    * Since boolean attributes only appear as positive conditions, each boolean attribute is assumed to be false unless noted as true. The value of multi-value discrete attributes taken as unknown ("?") unless a value is specified.

    * The original case identifications, p1 to p200 in the .data file and t1 to t26 in the .test file, have been added as a unique identifier attribute.

    [Note: in the original .data file, p165 has a repeated specification of o_ar_c(normal); p166 has repeated specification of speech(normal) and conflicting values air(moderate) and air(mild). No other problems with the original data were noted.]

    Attribute Information:

    age_gt_60: f, t.
    air(): mild,moderate,severe,normal,profound.
    airBoneGap: f, t.
    ar_c(): normal,elevated,absent.
    ar_u(): normal,absent,elevated.
    bone(): mild,moderate,normal,unmeasured.
    boneAbnormal: f, t.
    bser(): normal,degraded.
    history_buzzing: f, t.
    history_dizziness: f, t.
    history_fluctuating: f, t.
    history_fullness: f, t.
    history_heredity: f, t.
    history_nausea: f, t.
    history_noise: f, t.
    history_recruitment: f, t.
    history_ringing: f, t.
    history_roaring: f, t.
    history_vomiting: f, t.
    late_wave_poor: f, t.
    m_at_2k: f, t.
    m_cond_lt_1k: f, t.
    m_gt_1k: f, t.
    m_m_gt_2k: f, t.
    m_m_sn: f, t.
    m_m_sn_gt_1k: f, t.
    m_m_sn_gt_2k: f, t.
    m_m_sn_gt_500: f, t.
    m_p_sn_gt_2k: f, t.
    m_s_gt_500: f, t.
    m_s_sn: f, t.
    m_s_sn_gt_1k: f, t.
    m_s_sn_gt_2k: f, t.
    m_s_sn_gt_3k: f, t.
    m_s_sn_gt_4k: f, t.
    m_sn_2_3k: f, t.
    m_sn_gt_1k: f, t.
    m_sn_gt_2k: f, t.
    m_sn_gt_3k: f, t.
    m_sn_gt_4k: f, t.
    m_sn_gt_500: f, t.
    m_sn_gt_6k: f, t.
    m_sn_lt_1k: f, t.
    m_sn_lt_2k: f, t.
    m_sn_lt_3k: f, t.
    middle_wave_poor: f, t.
    mod_gt_4k: f, t.
    mod_mixed: f, t.
    mod_s_mixed: f, t.
    mod_s_sn_gt_500: f, t.
    mod_sn: f, t.
    mod_sn_gt_1k: f, t.
    mod_sn_gt_2k: f, t.
    mod_sn_gt_3k: f, t.
    mod_sn_gt_4k: f, t.
    mod_sn_gt_500: f, t.
    notch_4k: f, t.
    notch_at_4k: f, t.
    o_ar_c(): normal,elevated,absent.
    o_ar_u(): normal,absent,elevated.
    s_sn_gt_1k: f, t.
    s_sn_gt_2k: f, t.
    s_sn_gt_4k: f, t.
    speech(): normal,good,very_good,very_poor,poor,unmeasured.
    static_normal: f, t.
    tymp(): a,as,b,ad,c.
    viith_nerve_signs: f, t.
    wave_V_delayed: f, t.
    waveform_ItoV_prolonged: f, t.
    indentifier (unique for each instance)


    Relevant Papers:

    Bareiss, E. Ray, & Porter, Bruce (1987). Protos: An Exemplar-Based Learning Apprentice. In the Proceedings of the 4th International Workshop on Machine Learning, 12-23, Irvine, CA: Morgan Kaufmann.
    [Web Link]




    • 分享你的想法


    所需积分:0 去赚积分?
    • 528浏览
    • 2下载
    • 0点赞
    • 收藏
    • 分享