Handling numeric attributes in Hoeffding trees

被引:0
|
作者
Pfahringer, Bernhard [1 ]
Holmes, Geoffrey [1 ]
Kirkby, Richard [1 ]
机构
[1] Univ Waikato, Hamilton, New Zealand
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
For conventional machine learning classification algorithms handling numeric attributes is relatively straightforward. Unsupervised and supervised solutions exist that either segment the data into pre-defined bins or sort the data and search for the best split points. Unfortunately, none of these solutions carry over particularly well to a data stream environment. Solutions for data streams have been proposed by several authors but as yet none have been compared empirically. In this paper we investigate a range of methods for multi-class tree-based classification where the handling of numeric attributes takes place as the tree is constructed. To this end, we extend an existing approximation approach, based on simple Gaussian approximation. We then compare this method with four approaches from the literature arriving at eight final algorithm configurations for testing. The solutions cover a range of options from perfectly accurate and memory intensive to highly approximate. All methods are tested using the Hoeffding tree classification algorithm. Surprisingly, the experimental comparison shows that the most approximate methods produce the most accurate trees by allowing for faster tree growth.
引用
收藏
页码:296 / 307
页数:12
相关论文
共 50 条
  • [1] Handling Numeric Behavioral Attributes in Actionable Behavioral Rules Mining
    Su, Peng
    Yang, Jian
    Li, Zhenpeng
    Liu, Yuan
    PROCEEDINGS 2016 IEEE INTERNATIONAL CONFERENCE ON SERVICE OPERATIONS AND LOGISTICS, AND INFORMATICS (SOLI), 2016, : 178 - 183
  • [2] Hoeffding Trees with nmin adaptation
    Garcia-Martin, Eva
    Lavesson, Niklas
    Grahn, Hakan
    Casalicchio, Emiliano
    Boeva, Veselka
    2018 IEEE 5TH INTERNATIONAL CONFERENCE ON DATA SCIENCE AND ADVANCED ANALYTICS (DSAA), 2018, : 70 - 79
  • [3] Ensembles of Restricted Hoeffding Trees
    Bifet, Albert
    Frank, Eibe
    Holmes, Geoff
    Pfahringer, Bernhard
    ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY, 2012, 3 (02)
  • [4] New options for Hoeffding trees
    Pfahringer, Bernhard
    Holmes, Geoffrey
    Kirkby, Richard
    AI 2007: ADVANCES IN ARTIFICIAL INTELLIGENCE, PROCEEDINGS, 2007, 4830 : 90 - 99
  • [5] Handling numeric attributes with ant colony based classifier for medical decision making
    Piculin, Matej
    Robnik-Sikonja, Marko
    EXPERT SYSTEMS WITH APPLICATIONS, 2014, 41 (16) : 7524 - 7535
  • [6] Analytical Split Value Calculation for Numerical Attributes in Hoeffding Trees with Misclassification-Based Impurity
    Mirkhan M.
    Amir Haeri M.
    Meybodi M.R.
    Annals of Data Science, 2021, 8 (03) : 645 - 665
  • [7] Induction of decision trees in numeric domains using set-valued attributes
    Kalles, Dimitrios
    Papagelis, Athanasios
    Ntoutsi, Eirini
    Intelligent Data Analysis, 2000, 4 (3-4) : 323 - 347
  • [8] Stress-testing Hoeffding trees
    Holmes, G
    Kirkby, R
    Pfahringer, B
    KNOWLEDGE DISCOVERY IN DATABASES: PKDD 2005, 2005, 3721 : 495 - 502
  • [9] Handling numeric attributes when comparing Bayesian network classifiers: does the discretization method matter?
    M. Julia Flores
    José A. Gámez
    Ana M. Martínez
    José M. Puerta
    Applied Intelligence, 2011, 34 : 372 - 385
  • [10] Handling numeric attributes when comparing Bayesian network classifiers: does the discretization method matter?
    Flores, M. Julia
    Gamez, Jose A.
    Martinez, Ana M.
    Puerta, Jose M.
    APPLIED INTELLIGENCE, 2011, 34 (03) : 372 - 385