Version space learning

Version space learning is a <a href="/facts/Symbolic_artificial_intelligence/ACBwbcG6">logical</a> approach to <a href="/facts/Machine_learning/e0w0XJTu">machine learning</a>, specifically <a href="/facts/Binary_classification/MeYYAMwp">binary classification</a>. Version space learning algorithms search a predefined space of <a href="/facts/Hypothesis/gMABEJul">hypotheses</a>, viewed as a set of <a href="/facts/Sentence_(logic)/lnsLzvTx">logical sentences</a>. Formally, the hypothesis space is a <a href="/facts/Logical_disjunction/jbI2EGlM">disjunction</a>

H
          
            1
          
        
        ∨
        
          H
          
            2
          
        
        ∨
        .
        .
        .
        ∨
        
          H
          
            n
          
        
      
    
    {\displaystyle H_{1}\lor H_{2}\lor ...\lor H_{n}}

(i.e., one or more of hypotheses 1 through n are true). A version space learning algorithm is presented with examples, which it will use to restrict its hypothesis space; for each example x, the hypotheses that are <a href="/facts/Consistency/IHkdnA9M">inconsistent</a> with x are removed from the space. This iterative refining of the hypothesis space is called the candidate elimination algorithm, the hypothesis space maintained inside the algorithm, its version space.

Version space learning open-in-new

Version space learning