【知识点】
Random Trees classifier、Boosting classifier, or MLP, or Knearest, or Nbayes, or Support Vector Machines
[Dataset] letter-recognition.data, 20000 * 17, the front row 16000 for training, 4000 lines tested.
1 as the first label, 16 as a characteristic value (an integer, is normalized to the range 0 to 15, and mainly statistical moments edge count), the following meanings:
1. Box Horizontal position of X-Box
2. Y- Vertical position of Box Box
3. Box width of width
4. High height of Box
5. The ON pixels onpix Total #
6. The bar Mean X-X of ON pixels in Box
7. The bar Mean Y-Y of ON pixels in Box
. 8. Mean X variance x2bar
9. The variance Y y2bar Mean
10. The XY Correlation Mean xybar
11. The x2ybr Mean of X * X * Y
12. The xy2br Mean of X * Y * Y
13. Mean Edge X-EGE COUNT left to right
14. A xegvy of X-EGE with Correlation Y
15. A Y-COUNT EGE Mean Edge Top to bottom
16. A yegvx Correlation of Y with X-EGE
