_PROBLEM CoEPrA-2006_Regression_001 _GROUP_NAME Gavin Cawley _GROUP_MEMBERS Gavin C. Cawley _ADDRESS School of Computing Sciences University of East Anglia Norwich, Norfolk, NR4 7TJ United Kingdom _MODELING_PROCEDURE The descriptors provided are first standardised, to have a mean of zero and unit variance. A least-squares support vector machine with a normalised linear kernel function was used. The regularisation parameters were set by minimising the leave-one-out cross-validation estimate of the sum-of-squares error using conjugate gradient decent. A nested leave-one-out cross-validation procedure was used for performance estimation, the test prediction represent the arithmetic mean of the predictions of the 89 models generated in the outermost leave-one-out cross-validation procedure. A variety of other approaches were investigated, using (nested) leave-one-out cross-validation to determine the best. The leave-one-out statistics for this model are: MSE = 0.279834 Q^2 = 0.720090 No feature selection was performed (feature selection has often been observed to reduce predictive accuracy, especially in the case of datasets with more features than patterns it is probably not a good idea). The regularisation built into LS-SVMs is the only method used to control over-fitting. _PREDICTION Obj_00001 5.220 Obj_00002 5.738 Obj_00003 4.845 Obj_00004 4.760 Obj_00005 7.231 Obj_00006 6.703 Obj_00007 5.381 Obj_00008 4.028 Obj_00009 5.598 Obj_00010 4.931 Obj_00011 6.167 Obj_00012 3.865 Obj_00013 5.717 Obj_00014 4.542 Obj_00015 6.517 Obj_00016 5.744 Obj_00017 4.805 Obj_00018 6.282 Obj_00019 4.416 Obj_00020 3.934 Obj_00021 4.094 Obj_00022 5.130 Obj_00023 4.768 Obj_00024 4.954 Obj_00025 4.926 Obj_00026 6.887 Obj_00027 6.239 Obj_00028 6.123 Obj_00029 4.901 Obj_00030 4.603 Obj_00031 5.488 Obj_00032 5.570 Obj_00033 5.592 Obj_00034 5.945 Obj_00035 5.679 Obj_00036 6.375 Obj_00037 4.322 Obj_00038 5.592 Obj_00039 5.215 Obj_00040 4.160 Obj_00041 5.877 Obj_00042 4.797 Obj_00043 7.127 Obj_00044 4.357 Obj_00045 5.605 Obj_00046 5.258 Obj_00047 4.076 Obj_00048 5.525 Obj_00049 4.770 Obj_00050 4.550 Obj_00051 5.723 Obj_00052 5.602 Obj_00053 5.395 Obj_00054 3.920 Obj_00055 4.967 Obj_00056 5.923 Obj_00057 4.274 Obj_00058 4.744 Obj_00059 7.264 Obj_00060 3.831 Obj_00061 5.660 Obj_00062 5.180 Obj_00063 5.101 Obj_00064 5.522 Obj_00065 5.776 Obj_00066 7.032 Obj_00067 6.225 Obj_00068 5.821 Obj_00069 5.747 Obj_00070 5.173 Obj_00071 3.842 Obj_00072 5.681 Obj_00073 4.931 Obj_00074 5.322 Obj_00075 6.020 Obj_00076 4.445 Obj_00077 6.273 Obj_00078 6.441 Obj_00079 4.173 Obj_00080 5.931 Obj_00081 5.842 Obj_00082 6.270 Obj_00083 4.485 Obj_00084 5.346 Obj_00085 6.954 Obj_00086 5.766 Obj_00087 5.394 Obj_00088 4.976