我正在尝试将一些样本分类为两个不同的类。每个样本包含几个功能。但问题是这些样本属于特定组,每个特征代表该特定组的某些属性。 通常我的数据如下所示:
+1 1:547.6949 3:-122.99 4:-150.7976 5:-39.5595 6:-39.31721
-1 1:165.0353 3:-108.6274 4:-155.5586 5:-25.6145 6:-24.43071
-1 1:232.7848 3:-134.1802 4:-157.5643 5:-39.3388 6:-39.202710
-1 1:299.3863 3:-125.8626 4:-150.9111 5:-49.781 6:-52.444
-1 1:161.2316 3:-120.7769 4:-167.4669 5:-26.4132 6:-26.4737
-1 1:334.6234 3:-129.6332 4:-148.9993 5:-58.6543 6:-61.318
-1 1:203.4685 3:-121.9352 4:-158.4352 5:-39.1206 6:-39.086
-1 1:206.0046 3:-126.0539 4:-158.7209 5:-41.1821 6:-43.845
-1 1:229.0727 3:-107.8596 4:-160.4539 5:-40.8166 6:-43.4803
-1 1:202.8949 3:-137.3044 4:-166.1187 5:-36.4276 6:-37.438
-1 1:361.9454 3:-88.4864 4:-160.604 5:-27.7027 6:-23.787
+1 1:491.9685 3:-98.0369 4:-155.9668 5:-37.3844 6:-33.716
-1 1:778.6312 3:-115.5249 4:-148.5153 5:-54.9574 6:-59.09651
-1 1:413.1565 3:-92.4645 4:-159.4521 5:-35.5019 6:-29.4924
-1 1:501.4577 3:-105.343 4:-156.6044 5:-39.2832 6:-43.42
-1 1:487.5412 3:-115.4553 4:-155.5021 5:-47.0739 6:-44.568
-1 1:445.5218 3:-109.6812 4:-157.8129 5:-51.4686 6:-45.212
-1 1:316.3106 3:-93.2688 4:-162.7669 5:-31.5366 6:-28.482
-1 1:284.8187 3:-95.0161 4:-154.7459 5:-37.6311 6:-25.738
-1 1:391.9559 3:-108.8012 4:-157.7027 5:-40.0802 6:-44.219
这些是两组数据的样本,每组只有一个(+1)类,其余为(-1)。每组具有相同大小的10个样本。我的问题是,如何使用考虑此分组的SVM分类器。或者如果SVM不是最佳选择,你有什么建议?