<br> 
<hr> 
<a name="tmva"></a> 
<h3>TMVA</h3>

    TMVA version 4.0.1 is included in this root release:

    <h4>Main changes and new features introduced with TMVA 4</h4>
    
    <ul>
      <li> Reorganisation of internal data handling and constructors
      of methods to allow to build arbitrary composite MVA methods,
      and to deal with multi-class classification and multi-target
      regression
        
      <li> Extended TMVA to multivariate multi-target regression
        
      <li> Any TMVA method can now be boosted (linearly or
      non-linearly)
        
      <li> Transformation of input variables can be chained as wished
        
      <li> Weight files are now in XML format
        
      <li> New MVA methods "PDE-Foam" and "LD", both featuring
      classification and regression
        
    </ul>


    <h5>Comments</h5>

    <p>
      <em>On XML format:</em><br>
      
      The old text format is obsolete though still readable in the
      application. Backward compatibility is NOT guaranteed. Please
      contact the authors if you require the reading of old text weight
      files in TMVA 4.
      
    <p>
      <em>Standard macros:</em><br>
      
      The structure of the standard macros has changed: macros are
      still in the "$ROOTSYS/tmva/test" directory, but distinguished for
      classification and regression examples:
      
    <pre>TMVAClassification.C, TMVAClassificationApplication.C TMVARegression.C, TMVARegressionApplication.C</pre>
    
    Classification and regression analysis (training) is analysed as
    usual via standard macros that can be called from dedicated
    GUIs.
    
    <p>
      <em>Regression:</em><br> 

    <ul>
      <li>Not yet available for all MVA methods. It exists for:
      PDE-RS, PDE-Foam, K-NN, LD, FDA, MLP, BDT for single targets
      (1D), and MLP for multiple targets (nD).
      
      <li>Not all transformation of input variables are available
      (only "Norm" so far).

      <li>Regression requires specific evaluation tools:

        <ul>
          <li> During the training we provide a ranking of input
            variables, using various criteria: correlations, transposed
            correlation, correlation ratio, and "mutual information" between
            input variables and regression target.  (Correlation ratio and
            mutual information implmentations provided by Moritz Backes,
            Geneva U)
            
          <li> After the training, the trained MVA methods are ranked wrt.
            the deviations between regression target and estimate.
            
          <li> Macros plot various deviation and correlation quantities.
            A new GUI (macros/TMVARegGui.C) collects these macros.
        </ul>
    </ul>


    <h4> Improvements of / new features for MVA methods </h4>
    <menu>
    <li> <em>Linear Discriminant:</em>
      Re-implementation of "Fisher" method as general linear discriminant ("LD"),
      which is also regression capable (so far: single-target only)

    <li> <em>PDEFoam:</em>
      PDE-Foam is a variation of the PDE-RS method using a self-adapting binning
      method to divide the multi-dimensional variable space into a finite number
      of hyper-rectangles (cells). The binning algorithm adjusts the size and
      position of a predefined number of cells such that the variance of the
      signal and background densities inside the cells reaches a minimum.

    <li> <em>BDT:</em>
      Introduced gradient boosting and stochastic gradient boosting for 
      classification with BDT (as desribed by Friedman 1999). See "BDTG" 
      example in TMVAClassification.C/cxx.

      A new option allows to restrict the maximum tree depth. This may be used to
      avoid overtraining and often gives better performance than pruning. (The
      pruning mechanism needs to be revisited)

    <li> <em>MLP:</em>
      Introduced recognition of convergence via general ConvergenceTest-class for
      interrupting computations when convergence is reached. This feature has is
      used now in MethodMLP. Improved treatment of event-weights in BFGS training.

      Implemented random and importance sampling of events in DataSet. Implemented
      the usage of this feature for MLP.
  
    <li> <em>TMlpANN</em> (interface to TMultiLayerPerceptron) now also uses event weights
      and writes standalone C++ class.

    <li> <em>k-NN:</em>
      A new global knn search function has been added to NodekNN that searches for
      k-nearest neighbor using event weights instead of raw event counts. ModulekNN
      has been modified to allow searches using "weight" or "count" option, where
      "count" is default. Added UseWeight option to MethodKNN to allow using of
      "weight" or "count". 
      (Work by Rustem Ospanov, CERN). 

    <li> <em>Likelihood</em> (and general PDF treatment):
      Adaptive smoothing the PDF class, allowing it to smooth between MinSmoothNum 
      (for regions with more signal) and MaxSmoothNum (for regions with less signal). 

      Configuration of the PDF parameters from the option string moved to PDF class,
      allowing the user to define all the PDF functionalities in every classifier
      the PDF is used (i.e., also for the MVA PDFs). The reading of these variables
      was removed from MethodBase and MethodLikelihood. This also allows improved 
      (full) PDF configuration of MVA output via the "CreateMvaPdf" option.
      (Work by Or Cohen, CERN & Weizmann)
 
    <li> <em>New generalisation methods</em>:

        <ul>
          <li> <em>MethodCompositeBase:</em> combines more than one
          classifier within one.

          <li> <em>MethodBoost:</em> boosts/bags any classifier
          type. A special booking procedure for it was added to
          Factory class.

          <li> <em>MethodDT:</em> a classifier composed of a single
          decision tree, boosted using MethodBoost. Results are
          compatible with BDT, but BDT remains the default for boosted
          decision trees, because it has pruning (among other
          additional features).

        </ul>
    </menu>

    <h5> Other improvements </h5>

    <ul>
      <li> Improved handling of small Likelihood values such that
      Likelihood performance increases in analyses with many variables
      (~>10). Thanks to Ralph Schaefer (Bonn U.) for reporting this.

      <li> Nicer plotting: custom variable titles and units can be
      assigned in "AddVariable" call.

      <li> Introduced the inverse transformation InverseTransform for
      the variable transformations into the framework. While this is
      not necessary for classification, it is necessary for
      regression. The inverse transformation of the normalization
      transformation has been implemented.

      <li> Started to extend the variable transformations to the
      regression targets as well.

      <li> MethodCuts now produces the 'optimal-cut' histograms needed
      by macro mvaeffs.C. (macro 5a of TMVAGui.C)
 
      <li> MsgLogger can be silenced in order to prevent excess output
      during boosting.

      <li> Third dataset type added centrally (Training, Validation
      and Testing). The validation data is split off the original
      training data set.

      <li> Update of GUI and other Macros according to the new
      features of PDF and the addition of MethodBoost.
    </ul>


    <h4>Updates in TMVA 4.0.1</h4>


    <ul>
      
      <li> "Spectator" variables can be defined now which are computed
      just as the input variables and which are written out into the
      TestTree, but which don't participate in any MVA calculation
      (useful for correlation studies).
        
      <li> New booking option "IgnoreNegWeightsInTraining" to test the
      effect of events with negative weights on the training. This is
      especially useful for methods, which do not properly deal with
      such events. Note that this new option is not available for all
      methods (a training interrupt is issued if not available).

    </ul>

    <h5>Bug fixes:</h5>
    <ul>
      <li> Fixed regression bug in VariableNormalizeTransform (Use
      number of targets from Event instead of DataSet)
            
      <li> Fixed Multitarget-Regression in PDEFoam, foam dimensions
      were miscalculated

      <li> Added writing of targets to the weight files in regression
      mode to fix problems in RegressionApplication

      <li> Added missing standard C++ header files missing to some
      classes, which lead to compilation failures on some
      architectures (thanks to Lucian Ancu, Nijmegen, for reporting
      these).

      <li> Added checks for unused options to Factory and
      DataSetFactory configuration options interpretation. Will now
      complain if wrong option labels are used.

      <li> Fixed standard creation of correlation matrix plots

      <li> Fixed internal mapping problem giving a fatal error message
      when destroying and recreating the Factory.

    </ul>