Changeset 4601:1f217594b15f in orange


Ignore:
Timestamp:
04/21/08 10:18:43 (6 years ago)
Author:
ales_erjavec <ales.erjavec@…>
Branch:
default
Convert:
2d2103069d0ffa71dade2b885141ce682f8fb027
Message:
 
File:
1 edited

Legend:

Unmodified
Added
Removed
  • orange/doc/reference/LinearLearner.htm

    r4598 r4601  
    77<index name="classifiers+logistic regresssion+linear"> 
    88<h1>Linear Learner</h1> 
    9 <p>orange.LinearLearner is a learner that uses the <a href="http://www.csie.ntu.edu.tw/~cjlin/liblinear/">LIBLINEAR library</a> as a backend. It is very fast on large datasets.</p> 
     9<p><code>orange.LinearLearner</code> is a learner that uses the <a href="http://www.csie.ntu.edu.tw/~cjlin/liblinear/">LIBLINEAR library</a> backend that is very fast on large datasets.</p> 
     10<index name="classifiers+logistic regresssion+linear learner"> 
    1011<h2>LinearLearner</h2> 
     12<p>Linear learner learnes the attribute weights using one of the four possible methods.</p> 
    1113<p class=section>Attributes</p> 
    1214<dl class=attributes> 
    1315  <dt>solver_type</dt> 
    1416  <dd>Specifiys whitch method to use. Can be one of the folowing: 
    15     <ul><li>orange.LinearLearner.L2_LR (L2-regularized logistic regression, default)</li> 
    16     <li>orange.LinearLearner.L2LOSS_SVM_DUAL</li> 
    17     <li>orange.LinearLearner.L2LOSS_SVM</li> 
    18     <li>orange.LinearLearner.L1LOSS_SVM_DUAL</li> 
     17    <ul><li><code>orange.LinearLearner.L2_LR (L2-regularized logistic regression, default)</li> 
     18    <li><code>orange.LinearLearner.L2LOSS_SVM_DUAL</code></li> 
     19    <li><code>orange.LinearLearner.L2LOSS_SVM</code></li> 
     20    <li><code>orange.LinearLearner.L1LOSS_SVM_DUAL</code></li> 
    1921    </ul> 
    20     Note that only L2_LR supports probabilty esstimations.</dd> 
     22    Note that only <code>L2_LR</code> supports probabilty esstimations.</dd> 
    2123  <dt>eps</dt> 
    2224  <dd>Stopping criteria (default 0.01)</dd> 
     
    2527</dl> 
    2628 
     29<index name="classifiers+logistic regresssion+linear classifier"> 
    2730<h2>LinearClassifeir</h2> 
     31<p>Linear classifiers that uses one class vs. rest strategy for multi-class classification. It supports probability esstimation only if it was build with L2-regularized logistic regression learner.</p> 
    2832<p class=section> Attributes</p> 
    2933<dl class=attributes> 
    3034  <dt>weights</dt> 
    31   <dd>computed variable weights</dd> 
     35  <dd>A list of computed weight vectors for all one class vs. rest classifiers</dd> 
    3236<dl> 
Note: See TracChangeset for help on using the changeset viewer.