_PROBLEM CoEPrA-2006_Regression_002

_GROUP_NAME Gavin Cawley

_GROUP_MEMBERS
Gavin Cawley

_ADDRESS
 School of Computing Sciences
 University of East Anglia
 Norwich NR4 7TJ, U.K.

_MODELLING PROCEDURE

A wide variety of approaches have been tried, with 
leave-one-out cross-validation used to select the best.  In
this case, a least-squares support vector machine with a
normalised linear kernel, acting on a 1-of-n encoding of the
amino acids.  Leave-one-out cross-valdiation is used to
optimise the regularisation parameter.  A committee of
LS-SVMs from the outer leave-one-out cross-valdiation is used
to form the predictions.  The leave-one-out statistics are:

MSE = 0.307659
Q^2 = 0.472426

The committee model is rather too large to send in one file,
but if anyone is really interested, I'll see what I can do to
sort it out somehow.

I am not all that confident that this model is doing anything
really useful and found this regression task rather noisy and
hard to model.

_PREDICTION
Obj_00001 7.682
Obj_00002 7.895
Obj_00003 7.919
Obj_00004 7.919
Obj_00005 7.895
Obj_00006 7.870
Obj_00007 8.085
Obj_00008 7.919
Obj_00009 6.824
Obj_00010 6.824
Obj_00011 8.146
Obj_00012 6.824
Obj_00013 7.590
Obj_00014 7.682
Obj_00015 7.919
Obj_00016 7.895
Obj_00017 7.990
Obj_00018 6.820
Obj_00019 6.647
Obj_00020 7.590
Obj_00021 6.647
Obj_00022 7.682
Obj_00023 7.895
Obj_00024 6.647
Obj_00025 7.590
Obj_00026 7.919
Obj_00027 7.590
Obj_00028 7.919
Obj_00029 8.085
Obj_00030 7.855
Obj_00031 6.647
Obj_00032 7.895
Obj_00033 7.590
Obj_00034 7.590
Obj_00035 6.824
Obj_00036 7.680
Obj_00037 8.070
Obj_00038 6.860
Obj_00039 7.919
Obj_00040 6.824
Obj_00041 6.824
Obj_00042 7.972
Obj_00043 8.070
Obj_00044 7.895
Obj_00045 6.647
Obj_00046 8.085
Obj_00047 5.734
Obj_00048 7.895
Obj_00049 8.161
Obj_00050 6.824
Obj_00051 8.085
Obj_00052 7.293
Obj_00053 6.824
Obj_00054 8.085
Obj_00055 7.682
Obj_00056 6.647
Obj_00057 7.919
Obj_00058 5.788
Obj_00059 7.590
Obj_00060 7.919
Obj_00061 7.758
Obj_00062 8.085
Obj_00063 8.085
Obj_00064 7.919
Obj_00065 6.088
Obj_00066 7.895
Obj_00067 7.895
Obj_00068 5.717
Obj_00069 6.937
Obj_00070 7.590
Obj_00071 8.070
Obj_00072 7.590
Obj_00073 8.070
Obj_00074 7.919
Obj_00075 8.360
Obj_00076 6.824