Xgboost is a boosting+decision trees toolkit, look at the microblog on the various Daniel said the effect is very good, so download one, use a bit, the installation steps are as follows.
The first step is to compile the build Xgboost.exe (for CLI) and Xgboost_wrapper.dll (for Python). With VS Open the Windows folder under the Xgboost-master source folder, open the solution, official note is need to use x64,release, but my computer is win 8.1 32 bit, had to choose Win 32
Select "Rebuild Solution", found an error, one of the errors is, "Error C3861:" "Sleep": "Cannot find the identifier ", the Internet search for a solution, originally sleep s must be uppercase, This seems to be a function in Windows, and then recompile the line, the other errors do not have to worry about, and finally in the Windows folder of the release folder generated the following file
The second step is to install the Xgboost Python library. In cmd, change the current directory to the Xgboost root file plus the Python-package folder below, and then run the Python setup.py install to complete the Xgboost installation. Direct use when using Xgboost
Import Xgboost as XGB
Note that when using Xgboost in Python, you need to indicate where the wrapper folder is located, such as using the following command
Sys.path.append (' C:\\.........\\xgboost\\wrapper ')
Since then, the Xgboost has been installed, should be able to use the normal
Resources
1,xgboost Official information, https://github.com/dmlc/xgboost/tree/master/windows
A blog post on the 2,csdn, with an accompanying example, http://blog.csdn.net/john159151/article/details/45549143
3,kaggle previous post, https://www.kaggle.com/c/otto-group-product-classification-challenge/forums/t/13043/ Run-xgboost-from-windows-and-python
Xgboost installation on Windows