Ex Parte Eder - Page 19

            Appeal 2007-2745                                                                                  
            Application 09/761,671                                                                            

        1       Bielinski does not characterize the time frame for analysis of future action, but             
        2   we find that such projected time frames typically include relatively short term time              
        3   frames because of the inherent uncertainty in projections that increases with time                
        4   frame.  We further find that there is nothing in Bielinski that would suggest that the            
        5   time frame for the projection phase of the analysis is incompatible with a shorter                
        6   time frame.                                                                                       
        7       The Appellant has not sustained its burden of showing the Examiner erred.                     
        8       (4) The Appellant argues that Rappaport’s use of a tree based model topology                  
        9   is incompatible with Brown’s network topology (Br. 13:Third ¶).                                   
       10       The Appellant has made a broad contention of the incompatibility of these                     
       11   methods without a specific showing of the nature of their incompatibility.  The                   
       12   Appellant bases this argument again on Rappaport rather than Bielinski as such,                   
       13   contending that Rappaport implicitly teaches a tree methodology.  We find that                    
       14   nothing in Rappaport specifically refers to a tree based model topology.  Rappaport               
       15   presents a figure of a tree diagram to represent the hierarchical nature of                       
       16   organizational costs and activities (FF 33), but makes no representation as to how                
       17   this is incorporated within the model.                                                            
       18       Even if Bielinski’s VBM were to employ a tree based methodology, we find                      
       19   nothing inconsistent with employing a neural network within each of the branches                  
       20   of the tree’s analysis.  Further, we find nothing incompatible with assigning neural              
       21   network analysis to Bielinski’s phase of finding driver candidates as in claim 69                 
       22   element [1.b.] and assigning a tree based induction model to identify drivers as in               
       23   element [1.c.].  The Appellant has not made any contention otherwise.                             
       24       The Appellant has not sustained its burden of showing the Examiner erred.                     


                                                      19                                                      


Page:  Previous  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  Next

Last modified: September 9, 2013